They wanted to give Google Ads a try.
When we first talked to the Jean Patrique owner, they wanted to give Google Ads a try, but they had never run it before. Until then, their primary source of conversions was Facebook Ads.
Google Ads is a more profound and complex platform to set up than Facebook, where the main worry is the creative. It is a thrilling challenge to plan the approach and the campaigns, and it is a way bigger investment.
Google Ads needs deep segmentation, audience research, keyword creation, and strategy. That was especially true back in 2019, when there were 40 segments to consider. Luckily the client gave us free rein to test new channels, and that is precisely what we did, first in the UK and soon after in the US.
More than 500 products to market well, twice over.
It was challenging, but rewarding in the end. One of the main impediments was that Jean Patrique is an ecommerce store with more than 500 products. The client found it increasingly difficult to market them all effectively.
Our job was to find the most effective way to advertise these products to the right audience, without going over the agreed ad spend, and to do it twice over: once for the UK, and once for a brand-new US store. To pull that off, we focused on three things:
- Conversion tracking
- Alpha-beta keyword structure
- Smart Shopping
2019 - Launch and the tracking foundation
The first year was about building two accounts from scratch. The UK account went live in the spring of 2019, the US account followed within weeks, and neither had history, tracking, or a product feed to lean on. We worked three things in sequence: honest conversion tracking, deep keyword research on a huge catalogue, and a disciplined alpha-beta Search build.
a Conversion tracking
Before running a single ad, we checked that every step inside the funnel was tracked correctly in Google Analytics and the Ads platform. We did not start the campaigns until we were sure the macro and micro conversions fired properly.
The client understood quickly that the number of sales alone was not enough data to work with in Google Ads. With micro conversions in place we could segment audiences and serve more relevant copy and banners in remarketing, and we always knew exactly where people fell out of the funnel.
Those micro events were also our early read on success before the real conversion volume arrived, which matters a lot when an account has no history. If your business is just starting with Google Ads, micro conversions are not optional, they are the first thing on the checklist.
A full macro and micro event layer in both markets, so every later decision ran on real signal.
b Alpha-beta keyword structure
At the same time we did the keyword research and built the structure. This is where our PPC, Web Analytics and CRO teams worked as one. We had no time to lose, so we planned the work to run in parallel.
The client already had a few search campaigns, so we analysed their search terms, put the converting ones into our structure, and built negative-keyword lists from the rest. We used an alpha-beta structure with the groups in SKAG. At first this let us push more budget into the research campaigns, the ones with proven broad-match keywords.
As those campaigns brought more conversions, we moved the converting search terms into exact-match Converting campaigns and put more budget behind them. With that structure we could grow Quality Score, sharpen the negatives to keep irrelevant traffic out, and lift the client's profitability. This is the kind of disciplined Google Ads management a 500-product catalogue needs from day one.
On bidding we did it properly, not by guessing. We started on manual CPC to gather impression-share and conversion-volume data, then tested enhanced CPC, Maximize Conversions, Target ROAS and Target CPA. We settled on Target CPA: it works over several weeks to collect data and tells you clearly when to adjust the target, or when to pause a keyword or an ad group.
An alpha-beta Search engine where the converting campaigns ran far more efficiently than the research ones, built to scale across 500+ products.
c Standard Shopping, built to read
We set up Shopping so the store could shine on Google, with every inch optimised for visibility. It started with keyword research for the products, to get the product titles right, because titles decide more than anything which searches trigger your product in the auction.
Our first move was Standard Shopping, where we could see exactly which search terms we were bidding on. We structured it by product type and by price range, and used Single Product Ad Groups (SPAGs), so we could see what each product triggered and adjust the negatives and the bids accordingly.
A Standard Shopping account structured to read every search term, the groundwork for the Smart Shopping migration that came next.
The founding year proved the model in both markets. Honest tracking, an alpha-beta Search build, and a readable Shopping account, all on two brand-new accounts with no history to lean on. Everything that followed was built on top of this.
We ran the measurement, not just the media.
From day one this was a full Web Analytics engagement, not only media buying, and that is a real part of why it worked. We owned the tracking: Google Analytics, Google Tag Manager, the conversion funnel, and the audiences, all built and maintained by us, in both markets.
Because we built the measurement before the spend, every optimisation since ran on signal we trusted. The events fired correctly, the values were right, and we could segment audiences by exactly where someone was in the funnel.
It also meant we could see the whole picture, not just the last click. When we mapped the assisted conversions, channels that looked quiet on last-click attribution turned out to be doing a lot of the work. Criteo retargeting drove a 2.27x assist-to-last-click ratio in the UK, doing far more in the middle of the funnel than a last-click report would ever admit, and email was carrying more of the journey than it ever got credit for.
Most agencies optimise to the last click and miss all of that. We fed it back into the bidding and the budget instead, which is a big part of why the returns held up as the account scaled.
2020 - The COVID breakout
2020 is the year everything we had built started compounding, helped along by a once-in-a-generation shift. When lockdowns hit, the country started cooking at home, and demand for quality kitchenware went through the roof. A well-built account was ready to catch it.
a The move to Smart Shopping
With enough conversion volume behind Standard Shopping, we began testing products on Smart Shopping. Once we saw the return, around 90% of the catalogue moved over. We kept working the feed the whole time, improving titles and restructuring by product profitability and sales volume.
The migration paid off directly: Shopping CPA fell 22% and Shopping ROAS rose 34% as Smart Shopping took over the heavy lifting, with the feed and the negatives still under our control.
This was also when the bidding matured. With real volume in the account we leaned on Target CPA and Target ROAS, and Target-CPA bidding cut CPA by 70% on one campaign type and 32% on another. The structure we had laid down in 2019 was finally running at full speed.
Around 90% of products moved to Smart Shopping, lifting Shopping ROAS 34% while cutting CPA 22%.
b Riding the surge, and a drop we caught
The lockdown surge was not a free lunch, it was a stress test, and a well-tracked account is what gets you through it. The brand even pivoted part of the range into Hygiene & Protection: hand gels, gloves, and protective gear for home and professional kitchens. Some of it we could run on Shopping, and some of it we could not. Google disapproved the face masks under its Shopping policy, so those sold organically while we leaned into the categories we were allowed to advertise.
Then in August the account wobbled. The data flagged it before anyone panicked: searches and clicks were up, but the conversion rate was sliding, which usually means lower-quality traffic creeping in. We diagnosed it with Google Trends and GA segmentation, then ran a hard negative-keyword sweep and pruned the weak ad groups.
Here is the twist we only saw because the tracking was honest: the users we acquired that August went on to have the highest transaction counts in September and October. The traffic was not bad, it was just badly attributed in the moment, and the long view proved it. That is the difference good measurement makes.
From there the year built toward its peak. Black Friday 2020 was a record by a distance. Against the brand's first, much smaller Black Friday in 2019, total revenue across all channels rose 742% and conversions 784%, and on Google Ads specifically revenue was up 860%. The Whatever Pan led the way, the hero product it always was.
The August quality drop, caught early and fixed: the traffic we kept delivered the highest transaction counts of the autumn.
Smart Shopping and a matured bidding stack drove the full-year climb: conversions +473% year on year, 2019 to 2020, at +62% higher profitability.
Black Friday 2020 set a record: all-channel revenue +742% and conversions +784% versus Black Friday 2019, with Google Ads revenue +860%.
2021 - Scaling efficiency across channels
By 2021 the account was not just bigger, it was sharper, and it had grown well beyond Google. The job shifted from proving the channel to squeezing more profit out of every part of it, across more channels than we started with.
a More conversions, better returns
The Search side kept compounding as the converting-keyword and negative lists matured. Year on year, 2021 versus 2020, Search non-brand delivered 108% more conversions at 139% higher ROAS, and Shopping 41% more conversions at 48% higher ROAS. That is the alpha-beta loop paying off: budget concentrating on what converts, waste flowing into the negatives.
The quality showed in the traffic, not just the totals. Non-brand impressions actually fell around 39% while clicks held roughly flat, which is exactly what you want to see: we were paying for fewer, more relevant impressions and a higher click-through rate, not buying cheap volume.
Across the whole account the year added up to +53% conversions and +42% total sales, at +29% higher profitability. Compounding, on an account that had started from zero two years earlier.
Non-brand impressions down around 39% while clicks held, so the account won better traffic, not cheaper traffic.
b Beyond Google: Bing, Facebook, and friends
We never treated Google as the whole world. By 2021 Jean Patrique was running a genuine multi-channel mix, each channel doing the job it was best at.
Bing punched well above its weight. It is cheaper and less competitive than Google, and on Jean Patrique it returned 6.57x on ad spend overall, with the brand campaigns up at 10.82x and higher. Facebook ran the top-of-funnel prospecting and the retargeting, especially around the Whatever Pan, the discovery engine that fed demand the search channels then captured.
Not everything worked, and we said so. We tested Pinterest properly, with a real keyword and interest plan, and it simply did not convert for this brand, so we stopped spending there rather than flatter the report. Criteo, on the other hand, earned its place as a mid-funnel workhorse, nurturing the people the prospecting brought in until they were ready to buy.
A real multi-channel mix: Google as the engine, Bing returning 6.57x, Facebook driving discovery, Criteo nurturing the middle, and Pinterest cut when it did not earn its keep.
c The Cooking Academy
Alongside the ads we helped feed the Cooking Academy, a recipe-led content engine tied into the email flows. Recipe pages became real steps on the path to a sale, pulling people in, warming them up, and handing them to the product pages. It is not a headline channel, but it is the kind of content-to-commerce flywheel that lowers how much you have to pay for the same customer.
One brand, two markets.
Jean Patrique is a UK business first, and the UK was the larger, broader engine: the full 500-product catalogue, brand awareness built over two decades, and the deepest channel mix. The US was the second account, launched within weeks of the UK in 2019.
The US was a tighter, more focused operation, built around the hero product. The Whatever Pan carried it, and it ran on Google and Bing rather than the full stack. At its peak the US account was around a quarter to a third the size of the UK, but per unit it held its own, at comparable returns, and on Bing the branded Whatever Pan shopping ran at exceptional ROAS.
Treating each market on its own terms, rather than as one blended number, is what kept both profitable. The UK got the breadth, the US got the focus, and the tracking on both meant we always knew which was which.
The channel they doubted became their biggest engine.
Read end to end, this account makes one argument. A brand that had only ever trusted Facebook handed us a channel they were unsure about, and we turned it into their biggest growth engine.
From a standing start in 2019, conversions grew +473% into 2020 and another +53% into 2021, with profitability up +62% and then +29% across those years. Total sales were up +42% in 2021 alone.
The growth was the good kind. Search non-brand returns rose +139% and Shopping +48% year on year, the Smart Shopping migration lifted Shopping ROAS 34% while cutting CPA 22%, and non-brand impressions fell while clicks held, so we were winning better traffic, not cheaper traffic.
And it was resilient. The account did not just ride the 2020 demand surge to a record Black Friday, up 742% in revenue against the year before, it survived the August quality wobble because the measurement caught it early. That is what a properly built account buys you: not just the good months, but a fast recovery from the bad ones.
Do not be afraid to try paid advertising.
They came to us leaning on Facebook. The channel they were unsure about, Google Ads, became their biggest growth engine, and it did it across two countries and half a dozen channels.
Jean Patrique is a textbook example of what an experienced PPC and Web Analytics team can do with a large catalogue: build the tracking honestly, structure the account so it can scale, and stay hands-on through the surges and the wobbles alike.
If you want to grow your ecommerce sales, do not be afraid of paid advertising. It can feel like a Sisyphus task at the start, with a 500-product feed and a blank account, but with the right team beside you the boulder stays at the top of the hill.
As an agency we specialise in helping stores like Jean Patrique grow their online profits through PPC, Web Analytics and CRO. We have years of experience, and we have seen, first hand, what works and what does not.
- Conversion tracking rebuilt in both markets (macro and micro events in GA and GTM)
- Alpha-beta Search structure with SKAG and disciplined negative lists
- Bidding tested to Target CPA, cutting CPA by 70%
- Smart Shopping migration: ROAS up 34%, CPA down 22%
- Multi-channel scale: Google, Bing (6.57x ROAS), Facebook and Criteo
- Attribution modelling that surfaced Criteo and email assist value
- Resilience through the 2020 surge and the August quality drop
