The revenue their dashboards couldn't see.

69% of this client's revenue arrives by phone or as manually created orders that Google Ads and GA4 cannot see at all. I worked out where that revenue really came from, using three independent methods that all agreed. Budget decisions stopped being made on wrong numbers, and the store is now pacing towards its best year ever.

Client
UK ecommerce retailer
Platform
Google Ads
Figures
Verified July 2026, rounded

Finding the invisible revenue

01
69%
of revenue was invisible to standard tracking

Most of this client's revenue never touches the checkout that advertising platforms measure. Customers ring up, or a member of staff keys the order in by hand. Google Ads and GA4 recorded none of it. On paper, advertising looked like a minor contributor.

The question that mattered: was PPC quietly driving any of that hidden revenue? I tested it three ways. A customer journey analysis, a cross-reference against the call log, and session data in BigQuery. All three methods converged on the same answer: PPC was originating roughly £4,000 to £5,000 a month of revenue that no report credited to it. Around £55,000 to £65,000 a year, once orders preceded by a call are included.

A £12,900 order from a repeat customer showed zero connection to advertising in every report. I traced its true origin to a PPC click using geographic segmentation, then built an offline conversion pipeline so that orders like it are credited automatically from then on.

The consequence is simple. The client's real return on ad spend is materially higher than any dashboard showed, and budget decisions are now based on evidence rather than whatever the platforms happened to be able to see.

Verified, year on year

02
£845k
2026 pacing. On track for the store's best year ever

Store revenue has grown every year since I took the account on. I won't claim the growth is all down to PPC; too many things move a business for that to be an honest claim. What I can verify is below, and every comparison is same-period, year on year.

£242k
2022*
£473k
2023
£596k
2024
£658k
2025
~£845k
2026†
Store revenue under my PPC management. *2022 partial year. †2026 pacing as at July.
~40%
True PPC-originated share of revenue, verified by multi-method attribution. Platform reports alone showed only around 28%.
£28k vs £600
In one month, a marketing platform's dashboard claimed £28,000 of revenue for this store; session-level evidence supported roughly £600. Platform-claimed attribution can overstate by 20 to 30 times. The figures on this page are the deflated, verified kind.
£55k-65k / yr
PPC-originated revenue recovered each year that no dashboard credited, now fed back into the account automatically.
3,600+
PPC-acquired customers in a lifetime-value tracking system covering £1.08m of cohort revenue. The average customer is worth around £298, about 1.3x their first order over 12 months, so bidding reflects long-term value rather than just the first sale.

Most agency fees grow as the account grows. Mine goes the other way. I aim to make myself effectively free: a fee covered by the performance improvements the work produces. Because I charge a percentage of PPC-attributed revenue with a fixed cap, cutting wasted spend cuts my own invoice, and every efficiency gain lands with the client. The fee has shrunk as a share of store revenue every year.

4.6%
2022
3.4%
2023
2.7%
2024
2.4%
2025
2.0%
2026

My fee as a share of total store revenue, by year. For scale: the invisible PPC-originated revenue recovered in the investigation above comes to roughly three times my monthly fee on this account.

Not just bids and budgets

03

Managing the platform is the smallest part of the job. The account has improved every year since 2022 because the digging never stops. Recent examples of work that sits outside the ads account:

2,300+ calls cross-referenced

Phone orders are invisible to every ad platform, so I cross-reference inbound calls against orders and ad clicks by phone number. In the first half of this year alone that covered over 2,300 calls, and found around £85,000 of revenue preceded by a call within 30 days of the order. That ongoing analysis is why proper call tracking is being added before the autumn peak, so those orders are credited at source.

20x return on one rerouted keyword

Quality Score problems are usually landing page problems. As the account expands into new search territory, I audit where every ad group sends its traffic and move it when the page doesn't match the search. One high-intent keyword converted at over 20x return on ad spend once routed to the right product page.

I also analyse where the site itself loses orders, from delivery-option defaults to missing category pages, and send costed proposals with the evidence attached rather than vague advice.

2,100+ customer chats mined

Sites drift. Prices change, options become paid extras, and ad copy that was true when written quietly stops being true. So I audit live ad claims against the current site on a rolling basis and rewrite anything that no longer holds, verifying each new claim against the live page before the old ad is paused. I have also mined over 2,100 customer chat conversations to find what buyers really ask before ordering, and fed that back into ad copy and page proposals.

Technologies on the account

04
Shopify
The store itself. Order and revenue data, the ground truth every figure is checked against.
Google PPC
The advertising platform, fed daily with offline conversions and competitor data, creating a custom dynamic bidding engine.
BigQuery
Session-level data warehouse used to verify what the platforms claim.
GA + offline measurement
Session analytics plus an offline conversion pipeline, so phone and manually created orders are credited to the clicks that drove them.
Custom scripts
Purpose-built code behind the daily monitors, tripwires and reporting below.
Cross-channel attribution
In-depth attribution analysis across channels, testing each platform's claims against independent evidence.
Daily monitors
Price accuracy and competitor auction positions checked every day, automatically.
Weekly CPA tripwire
Cost per acquisition thresholds with automatic failsafes to protect every penny spent.
Offline conversions
Daily uploads with enhanced matching, so phone and manual orders are credited.
Automated alerts
Oversight of every channel is built-in.
Status ● Running daily

Want numbers you can trust?

I take on a small number of PPC clients and measure my own work the same way I measured this. If that sounds like what you want, email me and we'll have a straightforward conversation. No pitch deck.

Talk to Pete [email protected]