173% More Paid Revenue From 5% Fewer Clicks
Stelrad is a UK radiator manufacturer selling a catalogue where height, finish and output decide the purchase. Absolute Digital Media rebuilt the Google Ads and Shopping account around product attributes rather than product ranges, and the most recent autumn season returned 13:1.
What did Absolute Digital Media achieve for Stelrad?
Absolute Digital Media's PPC campaign for Stelrad delivered +173% paid revenue (2021 to 2025), -5% clicks over the same period, +133% conversion rate (1.66% to 3.87%) and 13:1 return on ad spend (autumn 2025 season).
A technically differentiated catalogue in a seasonal category
Stelrad manufactures and supplies radiators, towel rails and heating products to homeowners, installers and merchants across the United Kingdom and Ireland, with group operations across mainland Europe. Stelrad Limited runs the UK business from its Mexborough site in South Yorkshire and is part of Stelrad Group plc, listed on the London Stock Exchange.
It sells a large and technically differentiated catalogue. A radiator is chosen on dimensions, output, finish and style rather than on brand alone, which means a customer arriving at the site usually knows roughly what they need and is selecting between variants rather than between manufacturers.
Demand is also strongly seasonal, concentrating through autumn and winter as heating becomes a priority, which compresses the window in which paid media has to perform. Stelrad has worked with Absolute Digital Media since January 2021 across paid media and digital PR.
Nearly three times the revenue from fewer clicks
All performance figures are taken directly from the Stelrad Google Ads account, extracted on 25 August 2026 and reported in the account currency. Paid revenue is the account’s conversion value; conversion rate is conversions divided by clicks; return on ad spend is conversion value divided by cost. Absolute spend and revenue figures are held back pending client approval, so each measure is shown as its reported movement, ratio or an index. The 2026 figures are year to date; annual comparisons use complete calendar years only. Semrush figures are the UK database’s modelled estimates of text-ad keyword coverage and are used as third-party evidence of visibility, not as a measure of the account.
A multi-award-winning agency
Integrated Agency of the Year - SilverEuropean Agency Awards 2026
Best Use of Search B2C (PPC) Small (The Swallow Co.)European Search Awards 2026
Standout B2C Ecommerce Website of the Year (Stelrad)UK Digital Excellence Awards 2026
Best Integrated Paid Media CampaignUK Paid Media Awards 2026
Best New Business CampaignUK Agency Awards 2025
Best Use Of Search (Retail/Ecommerce) - SEOUK Search Awards 2024Integrated Agency of the Year - SilverEuropean Agency Awards 2026Best Use of Search B2C (PPC) Small (The Swallow Co.)European Search Awards 2026Standout B2C Ecommerce Website of the Year (Stelrad)UK Digital Excellence Awards 2026Best Integrated Paid Media CampaignUK Paid Media Awards 2026Best New Business CampaignUK Agency Awards 2025Best Use Of Search (Retail/Ecommerce) - SEOUK Search Awards 2024 The account was bidding on ranges when customers were buying attributes
The choices that shaped the rebuild
Segment by attribute rather than by product range
Product groupings were built around the attributes that drive the purchase decision, including radiator height, design finish and price band, rather than around the catalogue structure the business uses internally. That allowed bids to follow the variants customers actually chose. Mirroring the website’s category tree is the intuitive approach, and it organises the account around how the business thinks about its products rather than how customers select them.
Rank SKUs by conversion value, not by traffic
Google Analytics and Google Ads data were blended to identify top performers across traffic, conversion value and return, which is a different list from the one traffic alone produces. Budget was then allocated against that list. In a catalogue with wide price variation, the highest-traffic product and the highest-value product are rarely the same one, and an account that cannot tell them apart will fund the wrong one.
Use supplementary feeds to give the platform better signals
Supplementary product feeds were deployed to modify product titles, apply custom labels and enable bidding tailored to attributes such as finish, height and range. Underperforming products were removed from the feed entirely to concentrate spend. A default feed, left as the ecommerce platform exported it, describes the catalogue accurately and tells Google nothing about which products deserve budget.
Sequence lower-funnel before upper-funnel
The build started with high-converting lower-funnel keywords and product groups, then expanded into upper-funnel awareness campaigns once the performance base was established and ahead of the peak sales period. Launching full-funnel simultaneously builds awareness before the conversion machinery works, which produces demand the account cannot capture efficiently, and in a compressed seasonal window there is no time to recover the difference.
Who ran what
Strategy, feed and account structure, led by Stephen McAllister
Stephen set the strategy, blended the analytics and ads data to rank the catalogue by conversion value, designed the attribute-level segmentation around height, finish and price band, built the supplementary feeds carrying the custom labels and optimised titles, and sequenced the three-phase build against the season.
Campaign management and optimisation, led by Tiffany Shears
The negative keyword audits and match-type rebuild in phase one, management through the feed changes and the scaling phase, and the weekly optimisation run from live product-level dashboards.
Creative, led by Shane Fox
Creative for the full-funnel structure introduced in phase three, covering the remarketing and brand campaigns that ran alongside Shopping ahead of the season.
Three phases against the demand curve
Performance clean-up to recover budget, then feed intelligence with custom labels and dynamic titles, then strategic scaling into high-margin groups and top-tier finishes ahead of the autumn peak.
Who worked on this account
Tiffany Shears leads the PPC department at Absolute Digital Media, running paid media across Google Ads, Microsoft Ads, Meta and TikTok for lead generation and eCommerce clients. She specialises in…
Edwina Lambourdiere is a Senior PPC Account Manager at Absolute Digital Media, running paid media accounts across Google Ads, Microsoft Advertising, LinkedIn Ads and Meta Ads for UK and international…
Alex is Head of Content at Absolute Digital Media, overseeing content production and ideation for regulated-sector clients in iGaming, finance, and healthcare. He specialises in creating content that performs reliably…
Shane Fox is Lead Designer at Absolute Digital Media, responsible for the agency’s creative output and for the design of everything a client actually sees: campaign creative, brand systems, landing…
Fewer clicks, nearly three times the revenue
The pairing that matters is revenue against clicks. The account bought fewer clicks in 2025 than in 2021 and turned them into nearly three times the revenue, because conversion rate more than doubled from 1.66% to 3.87% and the average order rose by a quarter. Spend more than doubled over the same period and return on ad spend still rose from 8.7:1 to 11:1, which is the reverse of what normally happens when a Shopping account scales. Against 2020, the year before Absolute Digital Media took over the account, 2025 paid revenue was 6.7 times higher at a return that had not fallen.
The clearest single moment is October 2023, when the feed was pruned to the products that converted. Three quarters of the impressions and more than half of the spend were removed, and revenue held within 3%. That is what removing products from a feed looks like when it is done on conversion value rather than instinct: the impressions that went were not producing orders, and the budget they were consuming was returned to the business. Return on ad spend more than doubled in a single quarter, ahead of the peak season rather than inside it.
The rebuild was not a one-off. In the twelve months to August 2026 the account delivered 35% more revenue than the twelve before it on 27% more spend at 10.3:1, and the autumn 2025 season delivered 63% more revenue than autumn 2024 on 19% more spend at 13:1. The rolling conversion rate dipped as the account expanded reach again in early 2026; the season figure shows the structure still does its job when demand arrives.
| Measure | 2021 | 2025 | Movement |
|---|---|---|---|
| Paid revenue | Index 100 | Index 273 | +173% |
| Conversions | Index 100 | Index 221 | +121% |
| Clicks | Index 100 | Index 95 | −5% |
| Conversion rate | 1.66% | 3.87% | +133% |
| Revenue per click | Index 100 | Index 288 | +188% |
| Average order value | Index 100 | Index 124 | +24% |
| Return on ad spend | 8.7:1 | 11.0:1 | +26% |
| Spend | Index 100 | Index 217 | +117% |
The October 2023 feed prune, two months either side: impressions index 100 to 24 (−76%), clicks 100 to 63 (−37%), spend 100 to 43 (−57%), paid revenue 100 to 97 (−3%), conversions 100 to 99 (−1%), conversion rate 3.33% to 5.22% (+57%), return on ad spend 6.6:1 to 14.8:1 (+124%), revenue per click 100 to 153 (+53%). Semrush's UK index shows the text-ad footprint expanding from 36 keywords in January 2023 to 558 in July 2023 as the account scaled ahead of the season, then falling away from October 2023 as the account concentrated on Shopping, which the index does not count as keywords; the account data shows paid activity continuing throughout, so the corroboration is used for the scale-up only.
What manufacturers and large-catalogue retailers should take from this
What Stelrad say
Ecommerce and Shopping PPC, answered
How should a large product catalogue be structured in Google Ads?
Around the attributes customers choose on, not the categories the business uses internally. Stelrad’s account was rebuilt around radiator height, finish and price band rather than product ranges, which allowed bids to follow the variants that actually converted rather than spreading budget evenly across the catalogue. Between the first year of the engagement and the latest full year, paid revenue rose 173% from 5% fewer clicks.
What is a supplementary product feed and why does it matter?
A supplementary feed adds or modifies data in your primary Shopping feed, allowing custom labels, optimised titles and attribute-level information to be applied without changing the ecommerce export. It is the mechanism by which Google is told which products deserve budget, and most accounts never build one.
Should you remove products from a Shopping feed?
Frequently, yes. Removing lines that do not convert concentrates budget on the ones that do, which matters most in seasonal categories where the selling window is short. It reduces the number of queries you can appear for and usually increases revenue.
How do you identify which products to prioritise in PPC?
By ranking them on conversion value and return rather than on traffic or click volume. In catalogues with wide price variation the highest-traffic product and the highest-value product are rarely the same, and blending analytics data with ads data is what separates them.
How can paid revenue rise while clicks fall?
By changing which clicks the account buys. Stelrad’s account delivered 173% more revenue in 2025 than in 2021 from 5% fewer clicks, because conversion rate more than doubled, from 1.7% to 3.9%, and revenue per click almost trebled. Structuring the account around the attributes customers choose on, and pruning the feed to the products that convert, replaces cheap unqualified traffic with expensive qualified traffic, and the second kind is worth far more per click.
Should you build full-funnel campaigns from the start?
Not usually. Establishing lower-funnel performance first means the awareness campaigns you add later feed a machine that converts. Building awareness before the conversion structure works generates demand the account captures inefficiently, and in a seasonal category there is no time to correct it.
How do you run PPC in a seasonal category?
By completing the structural work before the season rather than during it. Algorithms need learning periods, feeds need building and negative lists need auditing, none of which can be done at pace once demand ramps. The three-phase approach used for Stelrad front-loaded the clean-up and feed work so that scaling happened ahead of the autumn peak. In the autumn 2025 season the account delivered 63% more revenue than the season before on 19% more spend, at a 13:1 return.
How do you measure PPC success in ecommerce?
Against conversion value and return on ad spend, not traffic or conversion volume alone. For Stelrad the measures were paid revenue, up 173% between the first and latest full years of the engagement, and return on ad spend, which rose from 8.7:1 to 11:1 over the same period. Clicks describe activity. Revenue describes outcome.
More case studies
Spreading budget evenly across a big catalogue?
If you sell a large catalogue and your paid media spreads budget evenly across it, the conversation worth having is about which attributes your customers actually choose on and whether your feed knows about them.
