Audit of your shop · twelve months · example for a fictional shop

Meta fills the funnel,
brand search takes the credit

Brewline sells home espresso machines, grinders and beans from Rotterdam to five countries. This is one year of the shop read as a story, from the shop's own orders rather than from any advertising platform's dashboard: which ads really bring customers, what a customer is worth depending on what they bought first, which months each campaign made or lost money, how the two new channels (Bing, ChatGPT Ads) are doing, and where about €49,046 of profit a year is waiting.

Period 1 Oct 2025 to 30 Sep 2026Shop brewline.example, WooCommerce, NL · DE · BE · AT · FRSources orders, Google Ads, Meta, Microsoft Ads, ChatGPT Ads, GA4, Search Console, Klaviyo, Enalitica script, purchase pricesPrepared by Tilen Ledić, Enalitica
18 g in36 g out27 s93 °C9 barMill One · 2.4
Read this first

Five things to know, in sixty seconds

  1. Every euro spent on ads came back as €8.47 of sales and €2.80 of profit before the ad is paid. The business is healthy; the question is only where the next euro goes.
  2. Meta says it made €359,877. The orders can prove €114,115 that Meta finished and €253,434 that Meta touched. Meta is doing a different job than its dashboard claims: it starts journeys that Google and free search finish.
  3. €10,004 a year goes to ads on your own brand name. Most of those buyers had already decided; 24% of that revenue started with a Meta ad. Try the slider in section 4 to see what moving that money would do.
  4. What a customer buys first predicts what they are worth. A grinder buyer comes back 54% of the time and is worth 3.2× an accessory buyer after six months. The top 10% of customers bring 27% of the profit.
  5. Three weeks without the Mill One in November cost about €7,285 of profit, Black Friday's 15% on everything bought orders but not profit, and 1,832 abandoned carts never got a reminder. All three are fixes of an afternoon.
Sales per ad euro (ROAS)Sales divided by the advertising that brought them. A 5 means €5 of sales for every €1 of ads. It ignores what the product cost you, so it flatters low-margin sales.
Profit per ad euro (POAS)What is left after the product, delivery, fees and returns, divided by ad spend. Below 1.0 the ad loses money even when sales look fine. This is the number to budget on.
Sales per ad euro, whole shop (MER)All sales divided by all advertising, every channel together. It cannot be gamed by moving credit between campaigns, so it is the one figure to watch month after month.
Cost of a new customer (CAC)Advertising spent in a month divided by the customers who bought for the first time that month. Compare it with what a customer is worth, never with a single order.
What a customer is worth (LTV)The profit one customer brings over time: the first order plus every order after it. On this page it is shown after 3, 6 and 12 months, so you can see whether the cost of winning them was paid back.
Finished the sale / started the journeyEvery order is credited once to the channel the customer came from when they bought, and once to the channel that brought them for the first time. Both views are shown; neither is "the" truth on its own.
1 · The year in numbers

€1,267,148 of sales, €418,486 of profit, €149,571 of advertising

Sales are order totals including VAT, after the Black Friday discount. Profit is what is left after purchase prices, delivery the shop paid, payment fees and returns, before advertising.

Sales€1,267,1486,608 orders · average order €192
Profit before ads€418,48633.0% of sales · after returns and delivery
Advertising€149,571Google €92,436 · Meta €53,833 · Bing €2,316 · ChatGPT Ads €985
Sales and profit per ad euro8.5 · 2.80whole shop, all channels together
New customers4,1242,484 repeat orders · 37.6% of all orders
Returns3.0%199 orders · €77,152 refunded

From sales to profit, step by step

0k 400k 800k 1200k Sales: €1,267,148 €1,267k Sales Product cost: -€709,743 €710k Product cost Delivery paid by shop: -€27,965 €28k Delivery paid by Payment fees: -€26,592 €27k Payment fees Returns: -€8,342 €8k Returns Profit before ads: €418,486 €418k Profit before ads Advertising: -€149,571 €150k Advertising Profit after ads: €268,915 €269k Profit after ads
Where each euro of sales went. Product cost is the largest step, then advertising. Discounts are already inside the sales figure: Black Friday took €10,557 off list prices.

One euro of advertising, followed to the end

€1 spent on ads €1 spent on ads €8.47 of sales €8.47 of sales €2.80 of profit before ads €2.80 of profit before ads €1.80 kept after the ad is paid €1.80 kept after the ad is paid
For every €1 of ads the shop sold €8.47, kept €2.80 after product and delivery costs, and €1.80 after paying for the ad itself.

How Brewline compares with other European shops

Visitors who buy2.13%European median 2.66%
Email visitors who buy4.1%median 4.45%
Meta visitors who buy0.92%median 0.81%
Orders returned3.0%electronics about 11%, apparel 25%

Medians from a 2026 study of 117 European brands (486 million visits) and country guides; the shop's figures are its own orders divided by its own visits. Below the median on overall conversion and above it on email is the usual shape of a shop that sells considered purchases.

Sales and profit by month

salesprofit before ads
0k 50k 100k 150k Oct 25: sales €91,916, 341 orders Oct 25: profit €29,955 Oct 25 Nov 25: sales €133,536, 610 orders Nov 25: profit €38,228 Nov 25 Dec 25: sales €146,069, 611 orders Dec 25: profit €48,467 Dec 25 Jan 26: sales €89,746, 484 orders Jan 26: profit €29,691 Jan 26 Feb 26: sales €90,507, 472 orders Feb 26: profit €27,904 Feb 26 Mar 26: sales €105,687, 546 orders Mar 26: profit €33,889 Mar 26 Apr 26: sales €104,727, 518 orders Apr 26: profit €34,108 Apr 26 May 26: sales €101,541, 562 orders May 26: profit €34,652 May 26 Jun 26: sales €104,608, 568 orders Jun 26: profit €36,449 Jun 26 Jul 26: sales €88,348, 579 orders Jul 26: profit €31,917 Jul 26 Aug 26: sales €96,281, 617 orders Aug 26: profit €34,358 Aug 26 Sep 26: sales €114,182, 700 orders Sep 26: profit €38,868 Sep 26
November and December are 22% of the year. The profit bar shrinks more than the sales bar in November: the 15% Black Friday discount, more returns and the grinder that was out of stock (section 8).

Advertising by month

Google AdsMeta AdsMicrosoft Ads (from April)ChatGPT Ads (from July)
0k 5k 10k 15k 20k Oct 25: Google €7,091 Oct 25: Meta €4,223 Oct 25 Nov 25: Google €10,817 Nov 25: Meta €8,190 Nov 25 Dec 25: Google €10,226 Dec 25: Meta €6,349 Dec 25 Jan 26: Google €6,088 Jan 26: Meta €3,800 Jan 26 Feb 26: Google €5,657 Feb 26: Meta €3,454 Feb 26 Mar 26: Google €8,764 Mar 26: Meta €4,252 Mar 26 Apr 26: Google €7,734 Apr 26: Meta €4,083 Apr 26: Microsoft €382 Apr 26 May 26: Google €7,776 May 26: Meta €4,062 May 26: Microsoft €408 May 26 Jun 26: Google €6,622 Jun 26: Meta €3,574 Jun 26: Microsoft €347 Jun 26 Jul 26: Google €6,168 Jul 26: Meta €3,449 Jul 26: Microsoft €352 Jul 26: ChatGPT Ads €287 Jul 26 Aug 26: Google €6,997 Aug 26: Meta €3,673 Aug 26: Microsoft €378 Aug 26: ChatGPT Ads €310 Aug 26 Sep 26: Google €8,498 Sep 26: Meta €4,723 Sep 26: Microsoft €449 Sep 26: ChatGPT Ads €388 Sep 26
Meta doubled for Black Friday. Google stepped up in March, in one jump, and stayed there. Two new channels arrived in spring and summer; section 4 shows how they are doing.
MonthOrdersSalesAvg orderProfitAdsSales per ad €Profit per ad €NewRepeat
Oct 25341€91,916€270€29,955€11,3148.12.653410
Nov 25610€133,536€219€38,228€19,0077.02.0156941
Dec 25611€146,069€239€48,467€16,5758.82.92505106
Jan 26484€89,746€185€29,691€9,8889.13.00287197
Feb 26472€90,507€192€27,904€9,1129.93.06273199
Mar 26546€105,687€194€33,889€13,0168.12.60328218
Apr 26518€104,727€202€34,108€12,1998.62.80293225
May 26562€101,541€181€34,652€12,2468.32.83327235
Jun 26568€104,608€184€36,449€10,5429.93.46290278
Jul 26579€88,348€153€31,917€10,2568.63.11261318
Aug 26617€96,281€156€34,358€11,3598.53.02291326
Sep 26700€114,182€163€38,868€14,0578.12.76359341
Year6608€1,267,148€192€418,486€149,5718.52.8041242484
2 · Where the money came from

Free search finishes the most sales; Meta starts the most

One bar for the whole year. Switch between the channel the customer came from when they bought and the channel that brought them for the first time: the same money, two honest views.

Share of the year's €1,267,148 by channel

Organic: €287,822 (22.7%) Organic 23% Direct: €153,274 (12.1%) Direct 12% Google Ads: €483,056 (38.1%) Google 38% Meta Ads: €114,115 (9.0%) Meta 9% Microsoft Ads: €10,236 (0.8%) ChatGPT Ads: €3,032 (0.2%) Email: €65,623 (5.2%) Affiliate: €42,478 (3.4%) AI assistants: €26,640 (2.1%) Referral: €16,065 (1.3%) Unknown: €64,807 (5.1%)
Every order appears exactly once in each view. The grey slice is the 5.1% with no evidence at all; everything else has a proven source.

How a sale travels: from the channel that started it to the channel that finished it

Google Ads started, Google Ads finished: 101 orders (4%) Google Ads started, Meta Ads finished: 67 orders (3%) Google Ads started, Organic finished: 152 orders (6%) Google Ads started, Email finished: 149 orders (6%) Meta Ads started, Google Ads finished: 311 orders (13%) Meta Ads started, Meta Ads finished: 38 orders (2%) Meta Ads started, Microsoft Ads finished: 2 orders (0%) Meta Ads started, ChatGPT Ads finished: 1 orders (0%) Meta Ads started, Organic finished: 190 orders (8%) Meta Ads started, Direct finished: 144 orders (6%) Meta Ads started, Email finished: 69 orders (3%) Meta Ads started, Affiliate finished: 19 orders (1%) Meta Ads started, AI assistants finished: 14 orders (1%) Meta Ads started, Referral finished: 8 orders (0%) ChatGPT Ads started, Google Ads finished: 4 orders (0%) Organic started, Google Ads finished: 188 orders (8%) Organic started, Meta Ads finished: 54 orders (2%) Organic started, Microsoft Ads finished: 7 orders (0%) Organic started, ChatGPT Ads finished: 2 orders (0%) Organic started, Direct finished: 138 orders (6%) Organic started, Email finished: 89 orders (4%) Organic started, Affiliate finished: 29 orders (1%) Organic started, AI assistants finished: 15 orders (1%) Organic started, Referral finished: 5 orders (0%) Direct started, Google Ads finished: 26 orders (1%) Direct started, Organic finished: 25 orders (1%) Direct started, Email finished: 15 orders (1%) Email started, Direct finished: 101 orders (4%) Affiliate started, Google Ads finished: 122 orders (5%) Affiliate started, Meta Ads finished: 9 orders (0%) Affiliate started, Organic finished: 46 orders (2%) Affiliate started, Direct finished: 55 orders (2%) AI assistants started, Google Ads finished: 42 orders (2%) AI assistants started, Meta Ads finished: 3 orders (0%) AI assistants started, ChatGPT Ads finished: 1 orders (0%) AI assistants started, Organic finished: 67 orders (3%) AI assistants started, Direct finished: 44 orders (2%) Referral started, Google Ads finished: 7 orders (0%) Referral started, Organic finished: 14 orders (1%) Google Ads · 469 Meta Ads · 796 ChatGPT Ads · 4 Organic · 527 Direct · 66 Email · 101 Affiliate · 232 AI assistants · 157 Referral · 21 Google Ads · 801 Meta Ads · 171 Microsoft Ads · 9 ChatGPT Ads · 4 Organic · 494 Direct · 482 Email · 322 Affiliate · 48 AI assistants · 29 Referral · 13 started the journeyfinished the sale
Only the 2373 orders that involved two or more channels (36% of all orders). Follow the Meta ribbons: most of them land on Google brand search, free search and direct visits. That is money Meta spent and other channels collected.
ChannelFinished the sale, ordersSalesShareStarted the journey, ordersSalesAd spendSales per ad €, finished / started
Organic1424€287,82222.7%1457€291,985··
Direct1237€153,27412.1%821€99,285··
Google Ads1921€483,05638.1%1589€375,131€92,4365.23 / 4.06
Meta Ads457€114,1159.0%1082€223,875€53,8332.12 / 4.16
Microsoft Ads35€10,2360.8%26€6,979€2,3164.42 / 3.01
ChatGPT Ads11€3,0320.2%11€2,589€9853.08 / 2.63
Email810€65,6235.2%589€52,517··
Affiliate199€42,4783.4%383€83,614··
AI assistants115€26,6402.1%243€50,675··
Referral79€16,0651.3%87€15,690··
Unknown320€64,8075.1%320€64,807··
Total6608€1,267,148100%6608€1,267,148€149,5718.5

The same flow as a table: started by (rows), finished by (columns)

Google Meta Bing GPT ads Organic Direct Email Affiliate AI Referral Google Ads started by Google Ads, finished by Google Ads: 101 orders 101 started by Google Ads, finished by Meta Ads: 67 orders 67 started by Google Ads, finished by Microsoft Ads: 0 orders started by Google Ads, finished by ChatGPT Ads: 0 orders started by Google Ads, finished by Organic: 152 orders 152 started by Google Ads, finished by Direct: 0 orders started by Google Ads, finished by Email: 149 orders 149 started by Google Ads, finished by Affiliate: 0 orders started by Google Ads, finished by AI assistants: 0 orders started by Google Ads, finished by Referral: 0 orders Meta Ads started by Meta Ads, finished by Google Ads: 311 orders 311 started by Meta Ads, finished by Meta Ads: 38 orders 38 started by Meta Ads, finished by Microsoft Ads: 2 orders 2 started by Meta Ads, finished by ChatGPT Ads: 1 orders 1 started by Meta Ads, finished by Organic: 190 orders 190 started by Meta Ads, finished by Direct: 144 orders 144 started by Meta Ads, finished by Email: 69 orders 69 started by Meta Ads, finished by Affiliate: 19 orders 19 started by Meta Ads, finished by AI assistants: 14 orders 14 started by Meta Ads, finished by Referral: 8 orders 8 Microsoft Ads started by Microsoft Ads, finished by Google Ads: 0 orders started by Microsoft Ads, finished by Meta Ads: 0 orders started by Microsoft Ads, finished by Microsoft Ads: 0 orders started by Microsoft Ads, finished by ChatGPT Ads: 0 orders started by Microsoft Ads, finished by Organic: 0 orders started by Microsoft Ads, finished by Direct: 0 orders started by Microsoft Ads, finished by Email: 0 orders started by Microsoft Ads, finished by Affiliate: 0 orders started by Microsoft Ads, finished by AI assistants: 0 orders started by Microsoft Ads, finished by Referral: 0 orders ChatGPT Ads started by ChatGPT Ads, finished by Google Ads: 4 orders 4 started by ChatGPT Ads, finished by Meta Ads: 0 orders started by ChatGPT Ads, finished by Microsoft Ads: 0 orders started by ChatGPT Ads, finished by ChatGPT Ads: 0 orders started by ChatGPT Ads, finished by Organic: 0 orders started by ChatGPT Ads, finished by Direct: 0 orders started by ChatGPT Ads, finished by Email: 0 orders started by ChatGPT Ads, finished by Affiliate: 0 orders started by ChatGPT Ads, finished by AI assistants: 0 orders started by ChatGPT Ads, finished by Referral: 0 orders Organic started by Organic, finished by Google Ads: 188 orders 188 started by Organic, finished by Meta Ads: 54 orders 54 started by Organic, finished by Microsoft Ads: 7 orders 7 started by Organic, finished by ChatGPT Ads: 2 orders 2 started by Organic, finished by Organic: 0 orders started by Organic, finished by Direct: 138 orders 138 started by Organic, finished by Email: 89 orders 89 started by Organic, finished by Affiliate: 29 orders 29 started by Organic, finished by AI assistants: 15 orders 15 started by Organic, finished by Referral: 5 orders 5 Direct started by Direct, finished by Google Ads: 26 orders 26 started by Direct, finished by Meta Ads: 0 orders started by Direct, finished by Microsoft Ads: 0 orders started by Direct, finished by ChatGPT Ads: 0 orders started by Direct, finished by Organic: 25 orders 25 started by Direct, finished by Direct: 0 orders started by Direct, finished by Email: 15 orders 15 started by Direct, finished by Affiliate: 0 orders started by Direct, finished by AI assistants: 0 orders started by Direct, finished by Referral: 0 orders Email started by Email, finished by Google Ads: 0 orders started by Email, finished by Meta Ads: 0 orders started by Email, finished by Microsoft Ads: 0 orders started by Email, finished by ChatGPT Ads: 0 orders started by Email, finished by Organic: 0 orders started by Email, finished by Direct: 101 orders 101 started by Email, finished by Email: 0 orders started by Email, finished by Affiliate: 0 orders started by Email, finished by AI assistants: 0 orders started by Email, finished by Referral: 0 orders Affiliate started by Affiliate, finished by Google Ads: 122 orders 122 started by Affiliate, finished by Meta Ads: 9 orders 9 started by Affiliate, finished by Microsoft Ads: 0 orders started by Affiliate, finished by ChatGPT Ads: 0 orders started by Affiliate, finished by Organic: 46 orders 46 started by Affiliate, finished by Direct: 55 orders 55 started by Affiliate, finished by Email: 0 orders started by Affiliate, finished by Affiliate: 0 orders started by Affiliate, finished by AI assistants: 0 orders started by Affiliate, finished by Referral: 0 orders AI assistants started by AI assistants, finished by Google Ads: 42 orders 42 started by AI assistants, finished by Meta Ads: 3 orders 3 started by AI assistants, finished by Microsoft Ads: 0 orders started by AI assistants, finished by ChatGPT Ads: 1 orders 1 started by AI assistants, finished by Organic: 67 orders 67 started by AI assistants, finished by Direct: 44 orders 44 started by AI assistants, finished by Email: 0 orders started by AI assistants, finished by Affiliate: 0 orders started by AI assistants, finished by AI assistants: 0 orders started by AI assistants, finished by Referral: 0 orders Referral started by Referral, finished by Google Ads: 7 orders 7 started by Referral, finished by Meta Ads: 0 orders started by Referral, finished by Microsoft Ads: 0 orders started by Referral, finished by ChatGPT Ads: 0 orders started by Referral, finished by Organic: 14 orders 14 started by Referral, finished by Direct: 0 orders started by Referral, finished by Email: 0 orders started by Referral, finished by Affiliate: 0 orders started by Referral, finished by AI assistants: 0 orders started by Referral, finished by Referral: 0 orders
Read the Google column from top to bottom: Google brand search is where other channels' work gets paid out.
What this means for you. Judging Meta by the sales it finishes (€114,115) would get it cut. Judging it by the sales it starts (€223,875) keeps it. Both views are on this page so the decision is made with both, not with the platform's own number.
3 · What the platforms claim and what happened

Meta reports €359,877; the orders can prove €114,115 to €253,434

Each platform counts the sales it credits itself, including people who only saw an ad and bought a week later for another reason. Enalitica counts an order for a platform only when a click or a recorded visit proves it. The spread between the four bars is the size of the argument in every budget meeting.

Meta Ads, €53,833 spent

Meta reports €359,877 · 6.7 per ad € Any touch on the journey €253,434 · 4.7 per ad € Started the journey €223,875 · 4.2 per ad € Finished the sale €114,115 · 2.1 per ad €
758 orders worth €139,319 started with a Meta ad and ended on google ads (311), organic (190), direct (144).

Google Ads, €92,436 spent

Google reports €661,827 · 7.2 per ad € Any touch on the journey €542,481 · 5.9 per ad € Started the journey €375,131 · 4.1 per ad € Finished the sale €483,056 · 5.2 per ad €
Google is closer to the truth, and flattered by its own brand campaign: €173,722 of what it finished is people searching for "brewline", and 24% of that started with a Meta ad.
Orders GA4 could see78%of sales · 23% of orders had no GA4 visit (consent declined, ad blockers, iPhone privacy)
Orders with a proven source95%of sales · server-side click records do not depend on cookie consent
Truly unknown5.1%320 orders · €64,807 · shown as their own row, never spread over the others

Does Meta move people who never click? Monthly Meta spend against searches for the brand name

0 500 1,000 1,500 2,000 Oct 25: Meta €4,223, 754 brand searchesOct 25 Nov 25: Meta €8,190, 1,148 brand searchesNov 25 Dec 25: Meta €6,349, 1,101 brand searchesDec 25 Jan 26: Meta €3,800, 841 brand searchesJan 26 Feb 26: Meta €3,454, 840 brand searchesFeb 26 Mar 26: Meta €4,252, 970 brand searchesMar 26 Apr 26: Meta €4,083, 987 brand searchesApr 26 May 26: Meta €4,062, 1,044 brand searchesMay 26 Jun 26: Meta €3,574, 969 brand searchesJun 26 Jul 26: Meta €3,449, 962 brand searchesJul 26 Aug 26: Meta €3,673, 1,060 brand searchesAug 26 Sep 26: Meta €4,723, 1,387 brand searchesSep 26 €0k Meta €5k Meta
Each dot is a month. The months with more Meta spend are the months with more people typing "brewline" into Google (correlation 0.46). That is movement, not proof: Black Friday and word of mouth move the same dots. It is the one visible trace of advertising nobody clicked.
What this means for you. Any report built on Google Analytics is dividing by 78% of your sales, and the missing part is not spread evenly: phone visitors from Meta go missing more often than desktop visitors from Google. So Meta looks worse than it is and Google looks better. Every number on this page uses the proven 95%.
4 · Winners and losers, month by month

Google product search won 9 of 12 months; Meta new audiences lost 10

Each cell is profit per ad euro for that campaign in that month, after returns. Green is comfortably profitable (1.5 or more), amber is around break-even, red is losing money, grey is not running yet. This table does not exist in Google or Meta: each platform only knows its own spend and its own claims, not the profit on the orders.

Profit per ad euro, campaign by month

1.5 and above1.0 to 1.5below 1.0not running
Oct 25 Nov 25 Dec 25 Jan 26 Feb 26 Mar 26 Apr 26 May 26 Jun 26 Jul 26 Aug 26 Sep 26 Google Shopping (PMax) Google Shopping (PMax), Oct 25: €1.27 profit per ad euro, spend €3,671 1.3 Google Shopping (PMax), Nov 25: €1.04 profit per ad euro, spend €5,398 1.0 Google Shopping (PMax), Dec 25: €1.23 profit per ad euro, spend €5,381 1.2 Google Shopping (PMax), Jan 26: €1.04 profit per ad euro, spend €3,031 1.0 Google Shopping (PMax), Feb 26: €0.60 profit per ad euro, spend €2,874 0.6 Google Shopping (PMax), Mar 26: €0.83 profit per ad euro, spend €4,745 0.8 Google Shopping (PMax), Apr 26: €0.85 profit per ad euro, spend €3,950 0.8 Google Shopping (PMax), May 26: €0.71 profit per ad euro, spend €4,463 0.7 Google Shopping (PMax), Jun 26: €1.07 profit per ad euro, spend €3,787 1.1 Google Shopping (PMax), Jul 26: €1.21 profit per ad euro, spend €3,395 1.2 Google Shopping (PMax), Aug 26: €0.79 profit per ad euro, spend €3,950 0.8 Google Shopping (PMax), Sep 26: €0.81 profit per ad euro, spend €4,971 0.8 Google product search Google product search, Oct 25: €1.48 profit per ad euro, spend €2,559 1.5 Google product search, Nov 25: €1.31 profit per ad euro, spend €4,412 1.3 Google product search, Dec 25: €1.64 profit per ad euro, spend €3,893 1.6 Google product search, Jan 26: €1.61 profit per ad euro, spend €2,295 1.6 Google product search, Feb 26: €1.71 profit per ad euro, spend €2,005 1.7 Google product search, Mar 26: €1.35 profit per ad euro, spend €2,577 1.3 Google product search, Apr 26: €0.98 profit per ad euro, spend €2,354 1.0 Google product search, May 26: €1.77 profit per ad euro, spend €2,487 1.8 Google product search, Jun 26: €1.99 profit per ad euro, spend €2,077 2.0 Google product search, Jul 26: €1.42 profit per ad euro, spend €1,983 1.4 Google product search, Aug 26: €1.76 profit per ad euro, spend €2,289 1.8 Google product search, Sep 26: €1.98 profit per ad euro, spend €2,663 2.0 Google brand search Google brand search, Oct 25: €5.75 profit per ad euro, spend €860 5.8 Google brand search, Nov 25: €5.45 profit per ad euro, spend €1,006 5.4 Google brand search, Dec 25: €7.67 profit per ad euro, spend €951 7.7 Google brand search, Jan 26: €4.13 profit per ad euro, spend €761 4.1 Google brand search, Feb 26: €5.14 profit per ad euro, spend €778 5.1 Google brand search, Mar 26: €5.42 profit per ad euro, spend €832 5.4 Google brand search, Apr 26: €7.14 profit per ad euro, spend €820 7.1 Google brand search, May 26: €5.77 profit per ad euro, spend €826 5.8 Google brand search, Jun 26: €7.68 profit per ad euro, spend €758 7.7 Google brand search, Jul 26: €4.44 profit per ad euro, spend €790 4.4 Google brand search, Aug 26: €5.67 profit per ad euro, spend €758 5.7 Google brand search, Sep 26: €6.03 profit per ad euro, spend €864 6.0 Meta new audiences Meta new audiences, Oct 25: €0.32 profit per ad euro, spend €3,147 0.3 Meta new audiences, Nov 25: €0.27 profit per ad euro, spend €6,480 0.3 Meta new audiences, Dec 25: €0.30 profit per ad euro, spend €4,751 0.3 Meta new audiences, Jan 26: €0.57 profit per ad euro, spend €2,809 0.6 Meta new audiences, Feb 26: €0.50 profit per ad euro, spend €2,595 0.5 Meta new audiences, Mar 26: €0.41 profit per ad euro, spend €3,217 0.4 Meta new audiences, Apr 26: €0.29 profit per ad euro, spend €3,079 0.3 Meta new audiences, May 26: €0.60 profit per ad euro, spend €3,020 0.6 Meta new audiences, Jun 26: €0.39 profit per ad euro, spend €2,639 0.4 Meta new audiences, Jul 26: €0.83 profit per ad euro, spend €2,548 0.8 Meta new audiences, Aug 26: €0.64 profit per ad euro, spend €2,720 0.6 Meta new audiences, Sep 26: €0.68 profit per ad euro, spend €3,566 0.7 Meta retargeting Meta retargeting, Oct 25: €0.17 profit per ad euro, spend €1,076 0.2 Meta retargeting, Nov 25: €0.72 profit per ad euro, spend €1,710 0.7 Meta retargeting, Dec 25: €1.38 profit per ad euro, spend €1,598 1.4 Meta retargeting, Jan 26: €1.31 profit per ad euro, spend €992 1.3 Meta retargeting, Feb 26: €1.92 profit per ad euro, spend €859 1.9 Meta retargeting, Mar 26: €1.09 profit per ad euro, spend €1,035 1.1 Meta retargeting, Apr 26: €2.03 profit per ad euro, spend €1,005 2.0 Meta retargeting, May 26: €1.50 profit per ad euro, spend €1,042 1.5 Meta retargeting, Jun 26: €2.26 profit per ad euro, spend €935 2.3 Meta retargeting, Jul 26: €0.86 profit per ad euro, spend €901 0.9 Meta retargeting, Aug 26: €1.19 profit per ad euro, spend €954 1.2 Meta retargeting, Sep 26: €0.85 profit per ad euro, spend €1,157 0.8 Microsoft Ads (Bing) Microsoft Ads (Bing), Oct 25: not running Microsoft Ads (Bing), Nov 25: not running Microsoft Ads (Bing), Dec 25: not running Microsoft Ads (Bing), Jan 26: not running Microsoft Ads (Bing), Feb 26: not running Microsoft Ads (Bing), Mar 26: not running Microsoft Ads (Bing), Apr 26: €2.38 profit per ad euro, spend €382 2.4 Microsoft Ads (Bing), May 26: €1.16 profit per ad euro, spend €408 1.2 Microsoft Ads (Bing), Jun 26: €1.19 profit per ad euro, spend €347 1.2 Microsoft Ads (Bing), Jul 26: €1.32 profit per ad euro, spend €352 1.3 Microsoft Ads (Bing), Aug 26: €1.40 profit per ad euro, spend €378 1.4 Microsoft Ads (Bing), Sep 26: €1.84 profit per ad euro, spend €449 1.8 ChatGPT Ads ChatGPT Ads, Oct 25: not running ChatGPT Ads, Nov 25: not running ChatGPT Ads, Dec 25: not running ChatGPT Ads, Jan 26: not running ChatGPT Ads, Feb 26: not running ChatGPT Ads, Mar 26: not running ChatGPT Ads, Apr 26: not running ChatGPT Ads, May 26: not running ChatGPT Ads, Jun 26: not running ChatGPT Ads, Jul 26: €1.12 profit per ad euro, spend €287 1.1 ChatGPT Ads, Aug 26: €0.00 profit per ad euro, spend €310 0.0 ChatGPT Ads, Sep 26: €0.76 profit per ad euro, spend €388 0.8
Brand search is green every month because it catches people who already decided. Meta new audiences is red by the sale it finishes and would be amber to green by the journeys it starts (section 3). Microsoft Ads arrived in April and has been quietly profitable; ChatGPT Ads has three months and no verdict yet.
MonthSalesvs previous monthBest campaign (brand-name ads excluded)Worst campaignBest-selling product
Oct 25€91,916·Google product search €1.48 per ad €Meta retargeting €0.17 per ad €Alto Due espresso machine €27,713
Nov 25€133,53645%Google product search €1.31 per ad €Meta new audiences €0.27 per ad €Alto Due espresso machine €48,685
Dec 25€146,0699%Google product search €1.64 per ad €Meta new audiences €0.30 per ad €Alto Due espresso machine €44,940
Jan 26€89,74639%Google product search €1.61 per ad €Meta new audiences €0.57 per ad €Alto Due espresso machine €23,219
Feb 26€90,5071%Meta retargeting €1.92 per ad €Meta new audiences €0.50 per ad €Alto Due espresso machine €28,462
Mar 26€105,68717%Google product search €1.35 per ad €Meta new audiences €0.41 per ad €Alto Due espresso machine €26,964
Apr 26€104,7271%Microsoft Ads (Bing) €2.38 per ad €Meta new audiences €0.29 per ad €Alto Due espresso machine €26,964
May 26€101,5413%Google product search €1.77 per ad €Meta new audiences €0.60 per ad €Alto Due espresso machine €23,219
Jun 26€104,6083%Meta retargeting €2.26 per ad €Meta new audiences €0.39 per ad €Alto Due espresso machine €34,454
Jul 26€88,34816%Google product search €1.42 per ad €Meta new audiences €0.83 per ad €Alto Due espresso machine €20,223
Aug 26€96,2819%Google product search €1.76 per ad €ChatGPT Ads €0.00 per ad €Alto Due espresso machine €26,964
Sep 26€114,18219%Google product search €1.98 per ad €Meta new audiences €0.68 per ad €Alto Due espresso machine €24,717

The two new channels

ChannelSinceSpendOrdersNew customersProfit per ad €
Microsoft Ads (Bing)April€2,31635281.56
ChatGPT AdsJuly€98511100.62

Bing buyers skew to desktop and machines, which is why the profit per ad euro is good on a small budget. ChatGPT Ads buyers look like the free AI traffic (section 11): grinders and guides. Both are judged on orders with a proven click, the same rule as Google and Meta.

Try it yourself: move money between campaigns

Ad spend, change per year
€0
Profit, change per year
€0
New customers, change
0

Assumptions, stated: cutting brand-name ads loses 15% of their orders (the rest move to the free result), and every extra 10% of Meta budget brings 7% more profit, because audiences get more expensive as you buy more of them. Both come from this year's figures and both are worth testing before trusting.

What this means for you. The worst campaign is the same one 10 months out of twelve. That is not a bad month, it is a structural result: the campaign is being judged on a job it does not do. The fix is in section 13, and it is not "cut it".
5 · Customers and what they are worth

A new customer costs €36 and pays that back in the month of purchase

Customers grouped by the month of their first order. Acquisition cost is that month's ad spend divided by that month's new customers. Value is the profit a customer has brought so far, after product and delivery costs, before ads.

Out of 100 new customers, 47 buy again within six months

comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months comes back within 6 months one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only one purchase only
Customers from October to March, followed for six months. The 47 who came back added 17% on top of what the whole group spent on its first order.

A few customers carry the year

0% 25% 50% 75% 100% top 10% of customers = 27% of profit top 20% = 47% 0% of customers 25% of customers 50% of customers 75% of customers 100% of customers
Customers sorted from most to least profitable. The top 10% bring 27% of the profit, the top 20% bring 47%. The dashed line is what "every customer equal" would look like. These are the people the win-back flow and the loyalty offer should be written for.

What the first purchase predicts

Grinder first Grinder first: €94.01 profit per customer after 6 months, 661 customers Grinder first: €74.03 profit in the first month €94 after 6 months · 54% buy again · 661 customers Beans first Beans first: €56.29 profit per customer after 6 months, 443 customers Beans first: €21.29 profit in the first month €56 after 6 months · 72% buy again · 443 customers Machine first Machine first: €174.37 profit per customer after 6 months, 728 customers Machine first: €161.44 profit in the first month €174 after 6 months · 39% buy again · 728 customers Accessory first Accessory first: €29.43 profit per customer after 6 months, 471 customers Accessory first: €21.10 profit in the first month €29 after 6 months · 24% buy again · 471 customers
Dark part of the bar is the first month, light part is what followed. A grinder buyer is worth 3.2× an accessory buyer after six months, and 54% come back, mostly for beans.

What the second purchase is, given the first

then machine then grinder then beans then accessory Machine first Machine first, then machine: 2% of second orders 2% Machine first, then grinder: 6% of second orders 6% Machine first, then beans: 72% of second orders 72% Machine first, then accessory: 20% of second orders 20% Grinder first Grinder first, then machine: 2% of second orders 2% Grinder first, then grinder: 5% of second orders 5% Grinder first, then beans: 71% of second orders 71% Grinder first, then accessory: 22% of second orders 22% Beans first Beans first, then machine: 3% of second orders 3% Beans first, then grinder: 4% of second orders 4% Beans first, then beans: 71% of second orders 71% Beans first, then accessory: 23% of second orders 23% Accessory first Accessory first, then machine: 2% of second orders 2% Accessory first, then grinder: 8% of second orders 8% Accessory first, then beans: 65% of second orders 65% Accessory first, then accessory: 25% of second orders 25%
Whatever people buy first, the second purchase is mostly beans: the machine or grinder is the door, the beans are the relationship. The welcome flow should sell beans, not accessories.

Who comes back, month by month: each row is one month's new customers

month 1 month 2 month 3 month 4 month 5 month 6 Oct 25 Oct 25 customers, month 1 after the first order: 12.0% ordered 12% Oct 25 customers, month 2 after the first order: 13.8% ordered 14% Oct 25 customers, month 3 after the first order: 16.7% ordered 17% Oct 25 customers, month 4 after the first order: 7.3% ordered 7% Oct 25 customers, month 5 after the first order: 10.0% ordered 10% Oct 25 customers, month 6 after the first order: 9.1% ordered 9% Nov 25 Nov 25 customers, month 1 after the first order: 10.4% ordered 10% Nov 25 customers, month 2 after the first order: 11.1% ordered 11% Nov 25 customers, month 3 after the first order: 11.1% ordered 11% Nov 25 customers, month 4 after the first order: 7.4% ordered 7% Nov 25 customers, month 5 after the first order: 6.9% ordered 7% Nov 25 customers, month 6 after the first order: 5.8% ordered 6% Dec 25 Dec 25 customers, month 1 after the first order: 15.2% ordered 15% Dec 25 customers, month 2 after the first order: 11.9% ordered 12% Dec 25 customers, month 3 after the first order: 13.5% ordered 13% Dec 25 customers, month 4 after the first order: 7.7% ordered 8% Dec 25 customers, month 5 after the first order: 7.7% ordered 8% Dec 25 customers, month 6 after the first order: 6.9% ordered 7% Jan 26 Jan 26 customers, month 1 after the first order: 17.8% ordered 18% Jan 26 customers, month 2 after the first order: 11.8% ordered 12% Jan 26 customers, month 3 after the first order: 16.4% ordered 16% Jan 26 customers, month 4 after the first order: 6.6% ordered 7% Jan 26 customers, month 5 after the first order: 7.3% ordered 7% Jan 26 customers, month 6 after the first order: 7.7% ordered 8% Feb 26 Feb 26 customers, month 1 after the first order: 14.7% ordered 15% Feb 26 customers, month 2 after the first order: 12.8% ordered 13% Feb 26 customers, month 3 after the first order: 12.1% ordered 12% Feb 26 customers, month 4 after the first order: 9.5% ordered 10% Feb 26 customers, month 5 after the first order: 9.2% ordered 9% Feb 26 customers, month 6 after the first order: 8.8% ordered 9% Mar 26 Mar 26 customers, month 1 after the first order: 10.4% ordered 10% Mar 26 customers, month 2 after the first order: 14.0% ordered 14% Mar 26 customers, month 3 after the first order: 12.8% ordered 13% Mar 26 customers, month 4 after the first order: 8.2% ordered 8% Mar 26 customers, month 5 after the first order: 5.8% ordered 6% Mar 26 customers, month 6 after the first order: 7.9% ordered 8% Apr 26 Apr 26 customers, month 1 after the first order: 13.3% ordered 13% Apr 26 customers, month 2 after the first order: 14.7% ordered 15% Apr 26 customers, month 3 after the first order: 15.0% ordered 15% Apr 26 customers, month 4 after the first order: 6.8% ordered 7% Apr 26 customers, month 5 after the first order: 8.5% ordered 9% May 26 May 26 customers, month 1 after the first order: 15.3% ordered 15% May 26 customers, month 2 after the first order: 14.4% ordered 14% May 26 customers, month 3 after the first order: 13.1% ordered 13% May 26 customers, month 4 after the first order: 8.0% ordered 8% Jun 26 Jun 26 customers, month 1 after the first order: 11.7% ordered 12% Jun 26 customers, month 2 after the first order: 11.7% ordered 12% Jun 26 customers, month 3 after the first order: 14.1% ordered 14% Jul 26 Jul 26 customers, month 1 after the first order: 14.9% ordered 15% Jul 26 customers, month 2 after the first order: 13.4% ordered 13% Aug 26 Aug 26 customers, month 1 after the first order: 13.4% ordered 13% Sep 26
Each cell is the share of that month's new customers who placed an order in the first, second, third month after they arrived. The Black Friday customers (Nov 25 row) come back less than the October ones: a 15% discount buys a different kind of customer. The drop after month three is the beans cycle ending for people who did not resubscribe.

How long until the second order

0 to 30 days: 225 customers 2250 to 30 days 31 to 60: 473 customers 47331 to 60 61 to 90: 334 customers 33461 to 90 91 to 120: 233 customers 23391 to 120 121 to 150: 117 customers 117121 to 150 over 150: 372 customers 372over 150
1,754 customers ordered a second time. The most common gap is one to two months, which is one bag of beans; a reminder in week six lands exactly when the bag runs out. The long tail past five months is the machine buyers coming back for a grinder.

Beans buyers: share ordering again, by month after the first order

0% 10% 20% 30% 40% month 1: 27.1% ordered again 27% month 1 month 2: 26.6% ordered again 27% month 2 month 3: 27.5% ordered again 28% month 3 month 4: 10.6% ordered again 11% month 4 month 5: 12.4% ordered again 12% month 5 month 6: 7.9% ordered again 8% month 6
Steady for three months, then a drop after month four. A customer who has not reordered by month four has changed supplier; the fourth month deserves its own mail.
First order inNew customersCost per customerProfit, first monthAfter 3 monthsAfter 6 monthsAfter 12 monthsBought again within 6 months
Oct 25341€33.18€87.85€99.64€108.19€122.8052.2%
Nov 25569€33.40€66.00€74.74€81.31·42.9%
Dec 25505€32.82€90.31€102.30€108.16·45.0%
Jan 26287€34.45€83.90€96.59€102.02·49.1%
Feb 26273€33.38€78.34€90.74€99.53·50.5%
Mar 26328€39.68€83.08€95.72€102.60·44.2%
Apr 26293€41.64€93.98€106.72···
May 26327€37.45€84.23€94.87···
Jun 26290€36.35€96.07€105.39···
Jul 26261€39.29€90.21····
Aug 26291€39.03€86.24····
Sep 26359€39.16€82.84····

Cost of a customer against the profit of their first month

cost per customerprofit, first month
€0 €50 €100 €150 €200 Oct 25: cost per customer €33.18, 341 new customers Oct 25: profit per customer in month 0 €87.85 Oct 25 Nov 25: cost per customer €33.40, 569 new customers Nov 25: profit per customer in month 0 €66.00 Nov 25 Dec 25: cost per customer €32.82, 505 new customers Dec 25: profit per customer in month 0 €90.31 Dec 25 Jan 26: cost per customer €34.45, 287 new customers Jan 26: profit per customer in month 0 €83.90 Jan 26 Feb 26: cost per customer €33.38, 273 new customers Feb 26: profit per customer in month 0 €78.34 Feb 26 Mar 26: cost per customer €39.68, 328 new customers Mar 26: profit per customer in month 0 €83.08 Mar 26 Apr 26: cost per customer €41.64, 293 new customers Apr 26: profit per customer in month 0 €93.98 Apr 26 May 26: cost per customer €37.45, 327 new customers May 26: profit per customer in month 0 €84.23 May 26 Jun 26: cost per customer €36.35, 290 new customers Jun 26: profit per customer in month 0 €96.07 Jun 26 Jul 26: cost per customer €39.29, 261 new customers Jul 26: profit per customer in month 0 €90.21 Jul 26 Aug 26: cost per customer €39.03, 291 new customers Aug 26: profit per customer in month 0 €86.24 Aug 26 Sep 26: cost per customer €39.16, 359 new customers Sep 26: profit per customer in month 0 €82.84 Sep 26
Every month the green bar is taller than the orange one: the ads pay for themselves before the customer has even come back. The October 2025 group is worth €122.80 after a year against €33.18 to acquire.
What this means for you. You can afford up to about €50 for a new customer and still make money within three months; today you pay €36. Meta's new-audience campaign costs €76 per customer it starts and €169 per customer it finishes; the budget target belongs between those two, nowhere near the €29 its own dashboard implies.
6 · How people buy

62% buy on their first visit; the rest take about 1.6 days to decide

6,608 recorded journeys, from the first visit to the order.

Bought on the first visit62%1230 needed two visits · 863 three · 397 four or more
Time to decide, when it takes more than one visit1.6 daystypical · average 3.4 days because of a long tail
Machine buyers think longest1.6 daysgrinder 1.6 · beans 1.4 · accessory 1.5
On a phone58%of orders · desktop orders are bigger: €189 against €194

When people buy: sales by weekday and hour, whole year

00 03 06 09 12 15 18 21 Mon Mon 00:00: €3,652 in 12 months Mon 01:00: €479 in 12 months Mon 02:00: €34 in 12 months Mon 03:00: €78 in 12 months Mon 04:00: €770 in 12 months Mon 05:00: €1,057 in 12 months Mon 06:00: €1,962 in 12 months Mon 07:00: €7,717 in 12 months Mon 08:00: €6,751 in 12 months Mon 09:00: €6,360 in 12 months Mon 10:00: €12,562 in 12 months Mon 11:00: €8,566 in 12 months Mon 12:00: €6,018 in 12 months Mon 13:00: €9,693 in 12 months Mon 14:00: €7,597 in 12 months Mon 15:00: €8,582 in 12 months Mon 16:00: €5,266 in 12 months Mon 17:00: €10,689 in 12 months Mon 18:00: €7,383 in 12 months Mon 19:00: €13,759 in 12 months Mon 20:00: €17,470 in 12 months Mon 21:00: €16,972 in 12 months Mon 22:00: €15,407 in 12 months Mon 23:00: €9,010 in 12 months Tue Tue 00:00: €2,068 in 12 months Tue 01:00: €242 in 12 months Tue 02:00: €822 in 12 months Tue 03:00: €608 in 12 months Tue 04:00: €1,196 in 12 months Tue 05:00: €504 in 12 months Tue 06:00: €2,415 in 12 months Tue 07:00: €3,558 in 12 months Tue 08:00: €6,852 in 12 months Tue 09:00: €7,161 in 12 months Tue 10:00: €13,487 in 12 months Tue 11:00: €9,147 in 12 months Tue 12:00: €6,430 in 12 months Tue 13:00: €10,546 in 12 months Tue 14:00: €8,873 in 12 months Tue 15:00: €7,214 in 12 months Tue 16:00: €5,535 in 12 months Tue 17:00: €8,290 in 12 months Tue 18:00: €9,252 in 12 months Tue 19:00: €7,383 in 12 months Tue 20:00: €14,281 in 12 months Tue 21:00: €14,014 in 12 months Tue 22:00: €14,144 in 12 months Tue 23:00: €5,826 in 12 months Wed Wed 00:00: €2,536 in 12 months Wed 01:00: €1,053 in 12 months Wed 02:00: €136 in 12 months Wed 03:00: €0 in 12 months Wed 04:00: €368 in 12 months Wed 05:00: €102 in 12 months Wed 06:00: €1,850 in 12 months Wed 07:00: €4,568 in 12 months Wed 08:00: €6,445 in 12 months Wed 09:00: €6,710 in 12 months Wed 10:00: €13,238 in 12 months Wed 11:00: €14,234 in 12 months Wed 12:00: €12,038 in 12 months Wed 13:00: €11,844 in 12 months Wed 14:00: €5,926 in 12 months Wed 15:00: €10,260 in 12 months Wed 16:00: €4,254 in 12 months Wed 17:00: €3,834 in 12 months Wed 18:00: €10,029 in 12 months Wed 19:00: €14,717 in 12 months Wed 20:00: €15,870 in 12 months Wed 21:00: €14,026 in 12 months Wed 22:00: €12,245 in 12 months Wed 23:00: €7,442 in 12 months Thu Thu 00:00: €4,599 in 12 months Thu 01:00: €1,168 in 12 months Thu 02:00: €224 in 12 months Thu 03:00: €597 in 12 months Thu 04:00: €1,262 in 12 months Thu 05:00: €298 in 12 months Thu 06:00: €2,110 in 12 months Thu 07:00: €5,349 in 12 months Thu 08:00: €6,888 in 12 months Thu 09:00: €5,627 in 12 months Thu 10:00: €10,417 in 12 months Thu 11:00: €9,782 in 12 months Thu 12:00: €13,586 in 12 months Thu 13:00: €9,675 in 12 months Thu 14:00: €5,546 in 12 months Thu 15:00: €7,059 in 12 months Thu 16:00: €10,991 in 12 months Thu 17:00: €8,218 in 12 months Thu 18:00: €11,477 in 12 months Thu 19:00: €13,152 in 12 months Thu 20:00: €18,498 in 12 months Thu 21:00: €20,095 in 12 months Thu 22:00: €12,268 in 12 months Thu 23:00: €9,130 in 12 months Fri Fri 00:00: €3,544 in 12 months Fri 01:00: €1,005 in 12 months Fri 02:00: €409 in 12 months Fri 03:00: €616 in 12 months Fri 04:00: €34 in 12 months Fri 05:00: €1,106 in 12 months Fri 06:00: €4,113 in 12 months Fri 07:00: €4,480 in 12 months Fri 08:00: €10,766 in 12 months Fri 09:00: €10,611 in 12 months Fri 10:00: €8,831 in 12 months Fri 11:00: €12,987 in 12 months Fri 12:00: €11,435 in 12 months Fri 13:00: €7,750 in 12 months Fri 14:00: €8,766 in 12 months Fri 15:00: €8,160 in 12 months Fri 16:00: €8,345 in 12 months Fri 17:00: €4,824 in 12 months Fri 18:00: €8,699 in 12 months Fri 19:00: €13,146 in 12 months Fri 20:00: €22,393 in 12 months Fri 21:00: €18,246 in 12 months Fri 22:00: €11,506 in 12 months Fri 23:00: €8,003 in 12 months Sat Sat 00:00: €2,573 in 12 months Sat 01:00: €2,473 in 12 months Sat 02:00: €1,481 in 12 months Sat 03:00: €526 in 12 months Sat 04:00: €863 in 12 months Sat 05:00: €172 in 12 months Sat 06:00: €2,043 in 12 months Sat 07:00: €4,135 in 12 months Sat 08:00: €6,702 in 12 months Sat 09:00: €5,541 in 12 months Sat 10:00: €9,162 in 12 months Sat 11:00: €9,150 in 12 months Sat 12:00: €5,969 in 12 months Sat 13:00: €10,709 in 12 months Sat 14:00: €10,125 in 12 months Sat 15:00: €6,031 in 12 months Sat 16:00: €6,854 in 12 months Sat 17:00: €9,216 in 12 months Sat 18:00: €11,989 in 12 months Sat 19:00: €12,906 in 12 months Sat 20:00: €15,953 in 12 months Sat 21:00: €11,992 in 12 months Sat 22:00: €14,206 in 12 months Sat 23:00: €9,467 in 12 months Sun Sun 00:00: €3,057 in 12 months Sun 01:00: €2,082 in 12 months Sun 02:00: €879 in 12 months Sun 03:00: €1,126 in 12 months Sun 04:00: €1,317 in 12 months Sun 05:00: €174 in 12 months Sun 06:00: €2,590 in 12 months Sun 07:00: €3,463 in 12 months Sun 08:00: €5,999 in 12 months Sun 09:00: €12,306 in 12 months Sun 10:00: €9,512 in 12 months Sun 11:00: €13,972 in 12 months Sun 12:00: €12,510 in 12 months Sun 13:00: €7,966 in 12 months Sun 14:00: €7,266 in 12 months Sun 15:00: €7,500 in 12 months Sun 16:00: €10,894 in 12 months Sun 17:00: €10,611 in 12 months Sun 18:00: €11,470 in 12 months Sun 19:00: €13,040 in 12 months Sun 20:00: €22,535 in 12 months Sun 21:00: €16,503 in 12 months Sun 22:00: €21,695 in 12 months Sun 23:00: €9,254 in 12 months
38% of sales land between 19:00 and midnight. Sunday is the biggest day (€207,720 for the year, 16%). Google and Meta spend the daily budget evenly across the day, so the morning gets the same money as the evening at a third of the buying rate.
What this means for you. A machine is a 2-day decision. An ad that is judged on same-day sales will always look like it failed for machines and succeeded for beans. The retargeting that runs in those 2 days is the cheapest sale in the account.
7 · What sells, and what earns

Machines are 62% of sales; grinders and beans are where the profit keeps coming

Ranked by profit, not by sales. The last four columns show which channel the buyers of each product came from.

ProductUnitsSalesProfitMarginUnits returnedGoogleMetaEmailOrganic
Alto Due espresso machine476€356,524€119,51134%3542%11%3%23%
Mill One burr grinder709€204,901€83,64041%2939%5%4%23%
Alto Uno espresso machine545€244,705€78,18632%4741%11%2%23%
Starter Set (Alto Uno + Mill Hand + scale)328€180,072€59,10033%2845%12%2%23%
House Blend 1 kg2896€98,464€50,89052%3724%5%17%22%
Mill Hand grinder600€53,400€25,28447%1234%5%5%27%
Single Origin 250 g2723€39,484€20,46752%3021%6%19%19%
Dose Scale601€35,459€18,56052%2132%6%8%22%

Products most often bought together

House Blend 1 kg + Single Origin 250 g 203 orders203 House Blend 1 kg + Descaler 3-pack 136 orders136 Single Origin 250 g + Descaler 3-pack 128 orders128 House Blend 1 kg + Mill One burr grinder 116 orders116 House Blend 1 kg + Mill Hand grinder 97 orders97 Alto Uno espresso machine + House Blend 1 kg 95 orders95 Single Origin 250 g + Mill One burr grinder 86 orders86
Pairs inside one order. Two bags of beans is the pair that sells itself, and beans with a grinder is the pair worth a bundle price. A machine with the descaler barely appears, although every machine needs one by month three: that is a bundle waiting to be made.

Your prices against three competitors, checked weekly

ProductBrewlineLowest competitorPosition
Alto Due espresso machine€749€739+1.4%
Alto Uno espresso machine€449€479-6.3%
Mill One burr grinder€289€259+11.6%
Mill Hand grinder€89€84+6.0%
Pour Kettle€79€69+14.5%
Dose Scale€59€54+9.3%
House Blend 1 kg€34·no match
Starter Set€549·no match

The Mill One is 11.6% above the lowest listed price and is also the product most often left in the cart. The Pour Kettle is 14.5% above and converts worst from Google Shopping. The Alto Uno is 6.3% below and could carry €20 more.

Category mix

Machines €781,301 (62%), grinders €258,301 (20%), beans €137,948 (11%), accessories €100,155 (8%). The Alto Uno and the Starter Set are the front door. The Mill One is the product whose buyers stay: a grinder buyer is worth €94 after six months.

8 · Stock, discounts, returns and delivery

Three weeks without the Mill One cost about €7,285 of profit; Black Friday bought orders, not profit

The operations side, read from the orders: what was not sold because it was not there, what the discount did, what came back, and what delivery does to the margin in each country.

Out of stock, 18 Nov to 9 Dec

The Mill One sold 2.7 units a day in the weeks around the gap. Over 22 days that is about 59 units, €17,118 of sales and €7,285 of profit that did not happen, in the best three weeks of the year. 98 visitors with a Mill One in the basket bought nothing or bought the cheaper Mill Hand.

The reorder point for November needs November's pace, not October's.

Black Friday, 21 to 30 November: 15% off everything

Orders per day31.7rest of November 14.7
Profit per order€44rest of November €83

Ten days, 317 orders, €59,822 of sales, €10,557 given away in discount, €13,922 of profit. 296 new customers came in; the table in section 5 shows they come back less often than October's. The discount bought volume and a weaker kind of customer. Next year: 15% on beans and accessories, full price on machines, which sell out in December anyway.

Share of orders returned, by channel

Google Ads Google Ads: 67 returns on 1921 orders 3.5% · 67 of 1921 orders Meta Ads Meta Ads: 27 returns on 457 orders 5.9% · 27 of 457 orders Organic Organic: 46 returns on 1424 orders 3.2% · 46 of 1424 orders Direct Direct: 17 returns on 1237 orders 1.4% · 17 of 1237 orders Email Email: 14 returns on 810 orders 1.7% · 14 of 810 orders Affiliate Affiliate: 6 returns on 199 orders 3.0% · 6 of 199 orders AI assistants AI assistants: 1 returns on 115 orders 0.9% · 1 of 115 orders Referral Referral: 2 returns on 79 orders 2.5% · 2 of 79 orders
A return costs delivery both ways plus restocking. Meta new-audience orders come back 3.1× as often as the shop average; the scoreboard in section 4 already prices that in.

Profit per order by country

CountryOrdersAvg orderDelivery paid by shopProfit per order
NL3392€193€2.52€65
DE1413€187€5.46€61
BE803€185€4.15€61
AT464€199€7.29€64
FR344€200€8.78€61
Other192€199€10.31€58

Free delivery above €75 means the shop pays €7.29 on an average Austrian order and €2.52 on a Dutch one. A beans-only order to Austria loses money on its own and only pays off if the customer comes back.

9 · Free traffic from Google

People searching for your name: up 61%. People searching for what you sell: down 23%

Clicks from Google's free results, twelve months against the twelve months before. Searches containing "brewline" against everything else. Free search still finishes 1424 orders and €287,822, but the ground under it is narrowing.

Monthly clicks from free Google results

product searches, this yearproduct searches, last yearbrand searches, this yearbrand searches, last year
0 3,000 6,000 9,000 Oct 25 Nov 25 Dec 25 Jan 26 Feb 26 Mar 26 Apr 26 May 26 Jun 26 Jul 26 Aug 26 Sep 26 product searches this year, Oct 25: 5,571 product searches this year, Nov 25: 7,741 product searches this year, Dec 25: 6,717 product searches this year, Jan 26: 4,759 product searches this year, Feb 26: 4,069 product searches this year, Mar 26: 4,686 product searches this year, Apr 26: 4,434 product searches this year, May 26: 4,059 product searches this year, Jun 26: 3,697 product searches this year, Jul 26: 3,431 product searches this year, Aug 26: 3,490 product searches this year, Sep 26: 3,639 3,639 product searches last year, Oct 25: 6,030 product searches last year, Nov 25: 8,463 product searches last year, Dec 25: 7,653 product searches last year, Jan 26: 5,727 product searches last year, Feb 26: 5,195 product searches last year, Mar 26: 6,019 product searches last year, Apr 26: 5,888 product searches last year, May 26: 5,805 product searches last year, Jun 26: 5,558 product searches last year, Jul 26: 5,265 product searches last year, Aug 26: 5,616 product searches last year, Sep 26: 6,130 6,130 brand searches this year, Oct 25: 754 brand searches this year, Nov 25: 1,148 brand searches this year, Dec 25: 1,101 brand searches this year, Jan 26: 841 brand searches this year, Feb 26: 840 brand searches this year, Mar 26: 970 brand searches this year, Apr 26: 987 brand searches this year, May 26: 1,044 brand searches this year, Jun 26: 969 brand searches this year, Jul 26: 962 brand searches this year, Aug 26: 1,060 brand searches this year, Sep 26: 1,387 1,387 brand searches last year, Oct 25: 634 brand searches last year, Nov 25: 809 brand searches last year, Dec 25: 753 brand searches last year, Jan 26: 589 brand searches last year, Feb 26: 566 brand searches last year, Mar 26: 616 brand searches last year, Apr 26: 589 brand searches last year, May 26: 597 brand searches last year, Jun 26: 572 brand searches last year, Jul 26: 547 brand searches last year, Aug 26: 558 brand searches last year, Sep 26: 661 661
Product searches: 73,349 clicks last year, 56,293 this year. Brand searches: 7,491 to 12,063. The brand line rises with Meta spend; the product line falls every month regardless of season.

Searches that lost the most clicks, September against September

20252026
best home espresso machine under 500 Sept 2025: 1840 clicks, position 3.2 Sept 2026: 740 clicks, position 6.8 1840740 burr grinder for espresso Sept 2025: 1510 clicks, position 2.4 Sept 2026: 610 clicks, position 5.9 1510610 espresso grinder under 300 Sept 2025: 980 clicks, position 2.1 Sept 2026: 520 clicks, position 3.4 980520 how to dial in espresso Sept 2025: 1260 clicks, position 4.0 Sept 2026: 870 clicks, position 5.2 1260870 pour over kettle Sept 2025: 640 clicks, position 3.1 Sept 2026: 230 clicks, position 8.7 640230 coffee scale with timer Sept 2025: 520 clicks, position 2.8 Sept 2026: 180 clicks, position 9.1 520180 descaling espresso machine Sept 2025: 470 clicks, position 5.5 Sept 2026: 330 clicks, position 7.0 470330 single origin coffee beans netherlands Sept 2025: 410 clicks, position 3.3 Sept 2026: 260 clicks, position 4.9 410260
"best home espresso machine under 500" slipped from position 3 to 7; "pour over kettle" from 3 to 9. The two guides that used to rank are two years old and the category pages carry no text. Meanwhile €10,004 of ads bought clicks on searches where the shop already ranks first for free.
What this means for you. The total organic number fell only 15%, which looks fine. Underneath, the free traffic that brings new customers halved and was replaced by people who already know you. Two refreshed guides and text on six category pages are the cheapest growth on this page.
10 · Google, read page by page

18,700 searches a month land on no page built for them; 6 of 9 categories carry no text

For the Audit we connected Google Search Console for a few weeks, read every search and every page, and crawled the shop the way Google does. This is what a search engine sees when it looks at Brewline: which searches have no home, which titles it shows and nobody clicks, which pages it has quietly dropped. The connection is closed and the data deleted when the Audit is handed over.

Searches with no landing page18,700a month, across five groups of long-tail searches
Clicks lost to weak titles4,971a year, pages ranking well but clicked below the curve
Categories without text6 of 983,600 impressions a month on them
Pages Google has dropped222of 1,240 in the sitemap

Searches that have no page to land on

Groups of searches people type, how many times a month Google showed Brewline for them, and where the click currently lands. In every row there are enough products for a category, but no category exists, so the visitor lands on a page that answers a different question and leaves.

Group of searches (sample of each)searchesshown a monthwhere the click lands todayavg. positionproducts that fit
espresso grinder under 300 / budget espresso grinder / best grinder for espresso 2026146,200/grinders (all grinders, no price filter)8.911
single origin beans / ethiopian coffee beans / light roast espresso beans114,100/beans (one page for all beans)7.39
espresso starter kit / beginner espresso setup / home espresso bundle93,700/espresso-machines/alto-uno (product page)9.83
descaling espresso machine / how to descale / descaler for espresso82,900/guides/dial-in-espresso (wrong guide)11.41
milk pitcher 500ml / barista pitcher / latte art jug61,800/accessories (whole category)6.14
What this means for you. A "grinders under €300" page, a "single origin" page and a "starter kit" page are three new categories built from products you already sell. At a cautious 3% click rate on 18,700 monthly impressions that is about 6,732 free visits a year that today go nowhere. The descaling searches need a short guide, not a category; the pitcher searches a filter.

Titles Google shows and nobody clicks

The ten most-shown pages: position against click rate

on or above the curvebelow the curve, title or description problemexpected click rate at that position
0% 4% 8% 12% 16% expected CTR at this position /espresso-machines: position 4.1, CTR 3.1%, expected 7.3% /espresso-machines /grinders: position 3.6, CTR 5.2%, expected 8.5% /grinders /grinders/mill-one: position 2.8, CTR 11.8%, expected 11.0% /grinders/mill-one /beans: position 6.2, CTR 2.2%, expected 4.6% /beans /espresso-machines/alto-uno: position 3.1, CTR 9.6%, expected 9.8% /guides/dial-in-espresso: position 5.2, CTR 3.4%, expected 5.8% /accessories/kettles: position 7.4, CTR 1.2%, expected 3.8% /accessories/scales: position 5.9, CTR 1.9%, expected 4.9% /guides/grinder-comparison: position 4.4, CTR 4.1%, expected 6.9% /beans/house-blend: position 2.2, CTR 14.1%, expected 14.0% position 1 position 2 position 3 position 4 position 5 position 6 position 7 position 8
A page at position 4 normally gets about 8% of clicks. The espresso machines category sits at 4.1 and gets 3.1%: its title is "Brewline | Shop", the same words as the home page. Bubble size is impressions. Hover for the figures.
Pageshown a monthpositionclick rateexpectedclicks lost a monthwhy
/espresso-machines48,2004.13.1%7.3%171generic title (shop name first, no product words)
/grinders36,9003.65.2%8.5%101title cut off in results (over 60 characters)
/beans19,8006.22.2%4.6%40no meta description, Google writes its own
/guides/dial-in-espresso15,3005.23.4%5.8%30title cut off in results (over 60 characters)
/accessories/kettles12,1007.41.2%3.8%26no meta description, Google writes its own
/accessories/scales9,8005.91.9%4.9%25generic title (shop name first, no product words)
/guides/grinder-comparison9,1004.44.1%6.9%21same title as another page
What this means for you. 414 clicks a month, 4,971 a year, from seven pages whose only problem is the line of text Google shows. New titles and descriptions for seven pages is an afternoon of work and needs no developer. Success: click rate on these pages at or above the curve within eight weeks.

Categories without a single sentence of text

Google ranks a category page on what is written on it. 6 of Brewline's 9 categories show only a grid of products, and those are the pages where product searches fell this year (section 9). The three categories with a short description (espresso machines, pitchers, starter sets) are also the three with the best click rate for their position.

Categoryproductstextshown a monthclicks
Espresso machines6yes48,2001,494
Grinders4none36,9001,919
Beans9none19,800436
Kettles3none12,100145
Scales5none9,800186
Tampers4none3,10062
Pitchers3yes2,60071
Descalers2none1,90038
Starter sets1yes5,400210

Fix: 80 to 150 words on each of the six, written for the buyer, not for Google: what the category is, how to choose, what the shop recommends. Grinders first, beans second, by impressions.

Pages Google has seen and dropped

Of the 1,240 addresses in Brewline's sitemap, Google keeps 1,018 in its index. The rest are not findable by anyone, whatever the page says.

Indexed: 1018 (82%) 1018 Crawled, not indexed: 142 (11%) 142 Duplicate without canonical: 61 (5%) Noindex by mistake: 19 (2%)
Indexed · 1018Crawled, not indexed · 142Duplicate without canonical · 61Noindex by mistake · 19

142 crawled, not indexed: colour and size variants of the same product, each on its own address with the same text. Google treats them as copies. One product page per product, variants as options on it.

61 duplicates without a canonical: the same page reachable with and without a trailing slash and through a "?sort=" filter. A one-line setting in the shop platform.

19 marked noindex by mistake: two whole categories (tampers, descalers) were set to "hide from search" during a redesign in March and never switched back. 5,000 impressions a month are still arriving on them from old links.

Languages: 41% of Dutch and German page pairs do not tell Google which is which, so German buyers are sometimes shown the Dutch page. One tag per page pair.

Product data Google reads: price and availability on 412 product pages

Google can show the price, stock state and rating straight in the search result when the page carries them in a form it can read (a "Product schema"). Where it is wrong, Google either shows nothing or shows the wrong thing, and the shop looks less trustworthy than a competitor whose price is visible.

Complete Product schema: 312 (76%) 312 Missing price: 38 (9%) 38 Wrong availability: 24 (6%) No schema at all: 38 (9%) 38
Complete Product schema · 312Missing price · 38Wrong availability · 24No schema at all · 38

38 pages send no price (the ones with a variant selector), 24 say "in stock" while the shop shows "sold out", among them the Mill One for the whole of November, so Google kept sending people to a product they could not buy. 38 pages, all the accessories added since June, carry no product data at all. Fix: one template change in the shop platform; we check the result page by page before the Audit closes.

Worth together, cautiously: about €8,821 of profit a year (€2,600 from the titles, €1,761 from three new categories at half the expected clicks, €4,461 from winning back half of the product searches lost against last year), none of it needing a single euro of advertising. It is the smallest number in section 14 and the one with the longest life: a title fixed once keeps paying.
11 · Three channels Google Analytics gets wrong

ChatGPT grew 7.8× in a year and converts 1.6× better than free search

AI assistants (GA4 files them under "Unassigned"), email flows without tags (GA4 files them under "Direct") and abandoned carts (GA4 never sees them at all).

AI assistants, free

165 visits in October 2025, 1,293 in September 2026, 8,579 for the year. 115 orders, €26,640. First orders from AI visitors are 47% grinders: the assistants cite the dial-in guide and the grinder comparison, not the home page. Those two pages are the ones to keep accurate and in stock. The paid ChatGPT Ads test (section 4) brings the same kind of buyer.

Email: the flows were selling all year, GA4 said "Direct"

The welcome, abandoned-cart, post-purchase and win-back mails went out without tracking tags until 2 June. Enalitica reads the click record on the order, so the flows were credited all year: €40,281 from flows and €25,342 from newsletters. GA4 saw €12,435 of email sales in the first eight months against €38,049 proven. In June GA4's email line doubled, and nothing about the email had changed.

Abandoned carts, twelve months

4,163 checkouts with an email typed and no order. 2,331 reminders sent (56%; the rest had not ticked the consent box). 312 orders came in after a reminder, €55,043. The 1,832 carts that got no reminder are the largest untouched pile on this page. European shops lose about 72% of carts; the question is not whether people leave, it is whether anyone writes to them.

Email sales by month: proven on orders against what GA4 reported

proven on ordersGA4 email channel
€0 €5,000 €10,000 €15,000 Oct 25 Nov 25 Dec 25 Jan 26 Feb 26 Mar 26 Apr 26 May 26 Jun 26 Jul 26 Aug 26 Sep 26 proven, Oct 25: €1,944 proven, Nov 25: €2,567 proven, Dec 25: €3,775 proven, Jan 26: €7,130 proven, Feb 26: €5,164 proven, Mar 26: €6,656 proven, Apr 26: €5,554 proven, May 26: €5,260 proven, Jun 26: €6,426 proven, Jul 26: €5,080 proven, Aug 26: €7,711 proven, Sep 26: €8,358 €8,358 GA4, Oct 25: €924 GA4, Nov 25: €1,256 GA4, Dec 25: €316 GA4, Jan 26: €1,968 GA4, Feb 26: €1,153 GA4, Mar 26: €2,763 GA4, Apr 26: €1,817 GA4, May 26: €2,237 GA4, Jun 26: €5,430 GA4, Jul 26: €3,703 GA4, Aug 26: €6,542 GA4, Sep 26: €5,804 €5,804
The gap closes in June when the flows got their tags. Any year-on-year email comparison in GA4 for 2026 will show growth that did not happen.
12 · The year's decisions, and whether they worked

Nine changes, judged on orders rather than on the platform's dashboard

Every budget move, pause, price and operational change found in the spend and order history, with equal windows before and after.

DateChangeAreaVerdictWhat the orders said
12 Nov 2025Meta new audiences to €95/day for Black Friday (+90%)Meta AdsworkingOrders that started with a Meta ad rose with the spend; brand search finished most of them. Judged on the whole journey, not on the last click.
18 Nov 2025Mill One out of stock until 9 DecStocknot working59 units of expected demand had nothing to buy; €17,118 of sales and €7,285 of profit did not happen in the best three weeks of the year.
21 Nov 2025Black Friday: 15% off everything for ten daysPricingmixedOrders per day 31.7 against 14.7 in the rest of November, but profit per order €44 against €83. The discount bought volume, not profit; 296 new customers came in, and they come back less often than October's.
3 Mar 2026Google Shopping budget +48% in one stepGoogle AdsmixedGoogle relearned; cost per order rose for two weeks before settling. A 20% step would have kept the history.
3 Mar 2026YouTube test, €610 a month, two monthsGoogle Adspaused okNo order with a proven click in eight weeks; paused in May. Sales did not miss it.
1 Apr 2026Microsoft Ads (Bing) switched on, about €430 a monthMicrosoft AdsworkingSmall and clean: €1.56 of profit per ad euro over 6 months, buyers skew to desktop and machines. Worth a second step of 20%.
2 Jun 2026Tracking tags added to all Klaviyo flowsEmailworkingEmail sales in GA4 doubled overnight. The flows had been selling all year; GA4 had filed them under Direct.
1 Jul 2026ChatGPT Ads test, about €360 a monthChatGPT Adstoo early11 orders in 3 months, €0.62 of profit per ad euro so far, half of them grinder buyers like the free AI traffic. Three more months before a verdict.
15 Aug 2026Free delivery threshold raised from €50 to €75OperationsworkingShare of orders under the threshold fell and the delivery subsidy per order dropped; the average order moved up with it.

March: Google Shopping budget +48% in one step

From €103 to €153 a day on 3 March. Spend per euro of sales went from €0.51 in February to €0.40 in March and €0.36 in April. Google relearned who to show the ads to before it settled. Steps of 20% keep what it already knows.

Meta new audiences: two honest numbers

By the sales it finishes, €0.45 of profit per ad euro. By the journeys it starts, about €1.71. Its orders also come back 9.4% of the time. The second number is the one to decide with; the first is the one Google Analytics shows.

YouTube: €1,220, no proven order

Two months, no order with a YouTube click or a recorded visit, and no rise in brand searches or direct orders in the same weeks. Paused in May; nothing moved. A test, correctly ended.

13 · What to fix

Ten things, in order, each with a number and a fix

Each carries a priority, the evidence, and a concrete fix with a success measure. Nothing changes without your decision.

High

€10,004 a year on your own brand name, paying for demand Meta and free search created

The brand campaign finishes 799 orders at 17.4 in sales per ad euro, but 24% of that money started with a Meta ad and most of the rest with a free search result where you already rank first. Pausing it will not lose the orders; it will move them one row down.

Fix: switch brand search off on weekdays for four weeks, keep weekends on. Success: brand-name orders (paid plus free) within 5% of before. Then move the saved €6,003 a year to Meta new audiences or Google product searches.
High

Free Google traffic for product searches down 23% in a year

Eight searches that fed the category pages lost half their clicks and three to five positions. The two guides that ranked are from 2024 and the category pages carry no text. Brand-name growth hides this in the total.

Fix: refresh the two guides (dial-in, grinder comparison; also what ChatGPT cites), write 150 words on each of the six category pages, link the guides from the product pages. Target: the eight searches back to 70% of last year's clicks by March.
Medium

18,700 searches a month with no page to land on, and seven titles Google shows that nobody clicks

Five groups of long-tail searches (grinders under €300, single origin beans, starter kits, descaling, pitchers) are shown 18,700 times a month and land on a page that answers a different question. Seven of the ten most-shown pages are clicked below what their position should earn, 4,971 clicks a year, because the title is the shop's name or is cut off. 19 pages are hidden from Google by a setting left over from the March redesign.

Fix: three new categories from existing products, new titles and descriptions on seven pages, six category texts, the noindex setting reversed, one product-data template fixed. No advertising budget. Success: product-search clicks (section 9) back to last year's level within six months; worth about €8,821 of profit a year.
High

The Mill One stock-out cost about €7,285 of profit in the best three weeks

2.7 units a day, 22 days, nothing to buy. Visitors with the grinder in the basket bought the cheaper Mill Hand or nothing, and grinder buyers are the most valuable customers the shop has.

Fix: reorder point for October and November set from November's pace and the supplier's lead time, with an alert at ten days of cover. A "notify me" button during any gap keeps the email.
Medium

Black Friday's 15% on everything bought orders and weaker customers

Orders per day 31.7 against 14.7, profit per order €44 against €83, and the November customers come back less often than October's. Machines would have sold in December at full price.

Fix: next year 15% on beans and accessories only, machines at full price with a free bag of beans. Judge it on profit per order and on the retention row, not on the sales total.
Medium

Meta new audiences is judged on the wrong job

By the sale it finishes it reads €1.65 in sales per ad euro; by the journeys it starts it reads €5.36. Its orders return 9.4% of the time. Neither Meta's own 6.7 nor the first figure is the one to budget on.

Fix: judge Meta new audiences on the new customers it starts and their six-month value; set the target at €50 per customer (section 5) and let the budget run to it. Enalitica's weekly campaign review reports against that target.
Medium

1,832 abandoned carts never got a reminder

Consent is the gate: 56% of carts had it. The reminder brings back 13.4% of the carts it reaches.

Fix: move the newsletter box above the payment button with one line on what the customer gets (the dial-in recipe for the machine in the cart). Target: reminders to 70% of carts. Expected: about €12,993 a year at the current rate.
Medium

Budget raised 48% in one step on Google Shopping

The March raise made Google relearn and the first fortnight cost more per order than February. The same money in steps of 20% would have kept the campaign's history and reached the same level in three weeks.

Fix: a 20% rule for every budget change, written into the weekly review. Enalitica's morning mail flags any larger step the next day.
Low

Ads spread evenly over a day that sells in the evening

38% of sales land after 19:00; Sunday is 16% of the week. Google and Meta budgets are spread evenly across hours and days.

Check: an ad schedule with +25% from 19:00 to 23:00 and on Sundays, −25% from 02:00 to 08:00, on Google product searches and Shopping, for four weeks. Compare cost per order by hour before and after.
Low

Austrian and French orders under €75 lose money on delivery

The shop pays €7.29 per Austrian order on average. A beans-only order to Austria earns less than its delivery costs.

Check: a €95 free-delivery threshold for AT and FR, or a beans subscription of three bags with delivery included.

What works and should be left alone

  • Every monthly group of new customers paid back its acquisition cost in the month of purchase; the October 2025 group is worth 3.7× its cost after a year.
  • 95% of sales have a proven source. The tracking is doing its job; only 5.1% is truly unknown.
  • Google product searches: €1.57 of profit per ad euro, 495 new customers, returns 3.5%. The cleanest campaign in the account.
  • Microsoft Ads: €1.56 of profit per ad euro on a small budget since April. Worth one 20% step.
  • Meta retargeting: green or amber every month on the scoreboard, the quiet performer.
  • Email flows carry €40,281 a year and are now tagged. The cart reminder brings back 13.4% of what it reaches.
  • The YouTube test was ended after two months with no proven order. That is how a test should end.
14 · What the next twelve months could add

About €49,046 of profit, from six changes that cost nothing to try

Worked out from this year's own figures, each with its assumption stated. Estimates, not promises; the point is the size and the order.

Brand-name ads€6,00360% of brand spend redirected, orders assumed to move to free search
Stock cover in November€7,285the Mill One profit that did not happen, assumed repeatable
Cart reminders to 70% of carts€12,993at this year's 13.4% recovery rate
Black Friday without machine discounts€6,188half the profit-per-order gap on the same number of orders
Evening and Sunday ad schedule€7,7568% better efficiency on two Google campaigns, the cautious end
Search fixes (section 10)€8,821titles, three new categories, half the lost product searches won back
15 · Ask your own numbers

Every figure on this page was one question to an assistant

This Audit was built by asking: Claude read the shop's reports, journeys and customers through Enalitica's AI access, read-only and without personal data. Click a question to see the answer the assistant gave. The same questions are yours to ask on any morning, in plain words.

Questions that made this page

Meta reports €359,877. On orders with a proven Meta click, Meta finished €114,115 of sales and touched €253,434. Meta started 758 orders worth €139,319 that other channels finished, mostly brand search and free search.
Of the €173,722 the brand campaign finished, 24% started with a Meta ad and most of the rest with a free search result. The brand campaign cost €10,004.
Grinder first: €94 of profit per customer after six months, 54% buy again. Machine first: €174, 39% buy again. Accessory first: €29.
About 59 units, €17,118 of sales and €7,285 of profit over 22 days, from a pace of 2.7 units a day in the weeks around the gap.
Ten days, 317 orders (31.7 a day against 14.7 in the rest of November), €10,557 given away, profit per order €44 against €83. 296 new customers, who come back less often than October's. Volume yes, profit no.
Meta new audiences, 10 of 12 months by the sales it finished. By the journeys it started it would be near break-even or better; that is the judging problem, not a campaign problem.
Five groups, 18,700 impressions a month. The largest: grinders under €300 and budget espresso grinders, 6,200 a month at position 8.9, landing on the full grinder category with no price filter; 11 products fit. Second: single origin and light roast beans, 4,100 a month, landing on the one page that holds all beans.
Two things changed. Six category pages have no text, so two-year-old guides carried the rankings and slipped. And in March the tampers and descalers categories were set to noindex during the redesign; 19 pages left Google's index and have not come back.
Microsoft Ads since April: €2,316 spent, 35 orders, €1.56 of profit per ad euro. ChatGPT Ads since July: €985, 11 orders, €0.62; too early for a verdict.

Suggested steps

StepWhoTimeEffect
Weekday pause of brand-name ads, 4 weeksBrewline10 minup to €6,003 a year freed
New titles and descriptions on seven pages, noindex reversedBrewlineone afternoon4,971 free clicks a year back
Reorder points from November's paceBrewline1 hourno repeat of the €7,285 gap
Consent box above the pay buttonBrewline (WooCommerce)30 minreminders from 56% to 70% of carts
Meta target: €50 per new customer it startsBrewline, Enalitica monitors15 minbudget judged on value, not on the platform's claim
Black Friday: discount beans and accessories, not machinesBrewlinedecisionprofit per order back to the November norm
Refresh two guides, text on six category pagesBrewline2 weekseight searches back to 70% of last year
Microsoft Ads +20%, ChatGPT Ads verdict in DecemberBrewline10 minthe two new channels judged on the same rule as the old ones
Ad schedule test, 4 weeksBrewline20 mincost per order by hour, before and after
20% rule for budget stepsboth·no more relearning weeks