Google Ads remarketing lets a business show ads to people who have already visited its website, used its app or shared their details, and it can be one of the most efficient uses of a pay-per-click budget. In most UK accounts we review, it is not. The typical setup is a single list called something like “All visitors – 30 days”, attached to one display campaign, showing the same banner to everyone who ever landed on the site: the buyer who abandoned a basket, the job applicant reading the careers page, the supplier checking an address, the existing customer logging in to pay an invoice and the visitor who left after three seconds.
That default treats very different people as if they were the same, and the budget follows the volume rather than the intent. Low-value visitors dominate a broad list, so most impressions go to people unlikely to buy, while the small group of high-intent visitors gets the same generic message as everyone else. Without frequency limits, the same people may see the same ad many times a day for weeks, which wastes money and can make the brand look desperate. And many sites fire remarketing tags before visitors have agreed to advertising cookies, which creates a compliance problem under UK rules on cookies and similar technologies.
This guide starts with why the one-list setup wastes budget, then covers audience list segmentation by intent and funnel stage, Customer Match targeting with first-party data, remarketing frequency capping, dynamic remarketing feeds for e-commerce and the UK consent requirements under the Privacy and Electronic Communications Regulations (PECR) that govern remarketing tags. It is written for business owners and marketing managers who run or oversee Google Ads, in-house or through an agency.
What Google Ads remarketing is, and what it is not
Remarketing, which Google also calls using “your data” segments, works by adding people to audience lists based on how they have interacted with your business, and then using those lists in campaigns. The most common sources are:
- Website visitors, added via the Google tag when they view pages or take actions on your site.
- App users, added based on in-app activity.
- YouTube viewers who have watched your videos or interacted with your channel.
- Customer Match lists, built from contact details your customers have provided to you, such as email addresses, uploaded securely to Google.
These lists can then be used to show ads across display, video, search, shopping, Demand Gen and Performance Max campaigns, either as targeting (only show to this list), as observation (show to everyone but report on and adjust for this list), as an audience signal that helps automated campaigns find similar people, or as an exclusion (never show to this list).
Remarketing is not a way to follow every visitor around the internet indefinitely, and it is not a substitute for a site that converts. If visitors leave because the site is slow, confusing or uncompetitive, remarketing pays to bring them back to the same problem. It works best as a way to give interested people a relevant reason to return, at the right moment, with a message that reflects what they were doing.
Before building anything new, add two exclusion lists to your existing remarketing: recent converters (people who bought or enquired in the last 30 days, or longer for considered purchases) and visitors to pages that indicate no buying intent, such as careers, supplier and login pages. It takes minutes and immediately stops a large share of wasted impressions.
Why one list showing one ad wastes budget
The problem with a single, undifferentiated remarketing list is not that it never works. It usually produces some conversions, which is why it survives. The problem is what it costs to produce them, and how much better a segmented approach performs with the same budget.
Low-intent visitors dominate the list
On most business sites, the majority of visitors leave quickly or view only one page. In a single list, they make up most of the audience, so they receive most of the impressions and most of the spend. Meanwhile, the visitors who viewed pricing, started a quote or added items to a basket are a small minority, receiving a small share of the budget despite being far more likely to convert.
Everyone gets the same message
A visitor who read one blog post needs a different message from one who abandoned a checkout. A generic brand ad is too weak for the second and too pushy for the first. Segmented lists let messages match where someone is in their decision.
Existing customers and non-buyers are included
Customers logging in, staff, job seekers and suppliers all visit business websites. Without exclusions, you pay to advertise to them.
Durations are rarely thought through
A 30-day window is a default, not a strategy. Someone who abandoned a basket an hour ago and someone who visited four weeks ago are in very different situations; a business with a six-month sales cycle may need much longer windows for some segments and much shorter for others.
Frequency runs unchecked
Without limits, a small number of people can see the same ad dozens of times. Beyond a point, each additional impression adds cost without adding conversions, and risks annoying people who might otherwise have returned.
UK remarketing accounts — the numbers
The figures below are indicative, drawn from Google Ads account audits we carry out for UK businesses. They are not a formal survey, but the pattern is consistent across sectors.
The second figure explains most of the waste. When seven in ten impressions go to the least engaged visitors, the campaign’s average performance is dragged down by people who were never likely to buy, and the strong performance of high-intent segments is hidden inside the average. Splitting the list usually reveals that a small segment is producing most of the conversions at a fraction of the cost.
One list versus segmented remarketing
The comparison shows the practical differences between the default setup and a segmented approach, using the same overall budget.
One all-visitors list
Default setup, one campaign, one message
Segmented by intent and stage
Several lists, matched messages, controlled exposure
Segmentation does not need to be elaborate. Four or five well-chosen lists, with sensible durations and exclusions, capture most of the benefit. The goal is to stop treating a basket abandoner and a three-second bounce as the same person.
Where single-list remarketing budget goes
The chart shows how remarketing spend typically divides across visitor types in single-list accounts we audit, before segmentation. Figures are indicative and vary by site.
Read the chart from the bottom up. The two highest-intent groups — abandoners and pricing or contact page viewers — typically receive less than a tenth of the spend, while visitors who bounced almost immediately receive close to half. Around one pound in six goes to people who should be excluded altogether: existing customers, recent converters, job seekers and suppliers. Segmentation reverses this, putting most of the budget where conversions are most likely. For a broader look at where Google Ads budgets leak, see our guide to Google Ads budget waste.
Audience list segmentation by intent and funnel stage
Good audience list segmentation starts with a simple question: what did this person do that tells us how close they are to buying? The answer suggests the list, the message, the duration and how much to spend reaching them.
High intent
People who added to a basket without buying, started but did not finish an enquiry or quote form, viewed pricing, or visited the contact page. These are the most valuable remarketing audiences and deserve the largest share of budget per person. Durations should usually be short — often seven to fourteen days for e-commerce, longer for considered B2B purchases — because intent fades. Messages can be direct: a reminder of the product, an answer to a likely objection, reassurance about delivery or guarantees, or a reason to act now where one genuinely exists.
Medium intent
People who viewed specific product, category or service pages, spent meaningful time on the site, or viewed several pages, but took no further step. They have shown interest in something specific. Messages should relate to what they viewed — the service, the product range, a relevant case study — and durations can be longer, perhaps thirty to ninety days depending on the sales cycle.
Low intent
People who read a blog post, landed on an information page or visited briefly. Many will never buy, but some are early in their research. Spend per person should be low, messages should be soft — useful content, brand introduction — and very short visits are often best excluded entirely. Some businesses choose not to remarket to this group at all, using the budget elsewhere.
Existing customers
Customers are a separate audience with separate goals: repeat purchase, cross-sell, renewal or referral. They should usually be excluded from acquisition remarketing and, where it makes sense, targeted in their own campaigns with appropriate messages. Customer Match, covered below, is often the best way to reach them.
Exclusions
Every remarketing campaign should exclude recent converters, so you do not pay to sell to someone who has just bought, and visitors to pages that indicate non-buying intent, such as careers, supplier, investor and account login pages. Depending on the business, very short visits may also be excluded.
Choosing durations
Membership duration is how long someone stays on a list after qualifying. Match it to how long the intent behind the list lasts. A good rule is to start with your typical time from first visit to purchase, then use shorter durations for high-intent lists and longer ones for medium-intent lists. Layering lists by recency — for example, basket abandoners in the last three days, and those from four to fourteen days — lets messages and bids differ for each.
Mind the minimum sizes
Google requires lists to reach a minimum number of active users before they can serve, and the threshold is higher for search than for display. For smaller sites, very finely sliced lists may never become usable. Start with a few broader segments and split them further only when volumes allow. Our guide to Google Ads audience targeting covers the wider range of audience types beyond your own data.
Scoring three remarketing segments
The score grid compares three common segments for a typical UK service business. A high badge means a high level of that characteristic.
The grid explains why budget should not be spread evenly. Quote form abandoners are few, likely to convert and quickly lose interest, so they justify higher spend per person over a short window with a specific message. Blog readers are numerous and unlikely to convert soon, so they justify low spend, soft messaging and tight frequency limits, because large lists of low-intent people are where fatigue and waste build up fastest.
What the default setup costs compared with segmentation
The table compares how a typical monthly remarketing budget of £2,000 might be spent under a single-list setup and a segmented one, for a UK service business. Figures are illustrative, based on patterns in accounts we have restructured; actual results vary by sector, site and offer.
| Segment | Single list: share of spend | Segmented: share of spend | Typical relative conversion rate |
|---|---|---|---|
| Abandoners and pricing or contact viewers | 8% (around £160) | 45% (around £900) | Highest |
| Service or product page viewers | 12% (around £240) | 30% (around £600) | Moderate |
| Blog and information readers | 18% (around £360) | 15% (around £300) | Low |
| Very short single-page visits | 46% (around £920) | Excluded or minimal | Very low |
| Customers, converters, job seekers, suppliers | 16% (around £320) | Excluded; customers in own campaign at 10% | Not applicable for acquisition |
In this illustration, the segmented approach moves roughly the segmented approach moves roughly £740 a month from people unlikely to buy into the segments most likely to convertpound;1,100 a month away from people unlikely to buy and into the two segments most likely to convert, with no change to the total budget. The exact gain depends on the business, but the direction is consistent: when spend follows intent rather than volume, cost per conversion usually falls. Accurate conversion tracking is essential to see the effect; our guide to setting up conversion tracking covers the foundations.
Remarketing maturity — where most UK accounts sit
Combining segmentation, exclusions, frequency control, first-party data use, consent handling and measurement gives an indication of overall remarketing maturity. The gauge reflects a typical UK SME account before any restructuring.
A score in the low thirties usually reflects an account where remarketing exists but has not been designed. There is one list or a handful of default ones, a single campaign, few exclusions, no deliberate frequency control, no Customer Match and, often, tags that are not properly governed by consent. Performance is reported as a single number that looks acceptable but conceals considerable waste.
Restructuring usually lifts an account into the sixties or seventies within a couple of months. The largest early gains come from exclusions and splitting high-intent visitors into their own lists; the later ones from Customer Match, dynamic remarketing where relevant, and measurement that tests whether remarketing is genuinely adding conversions.
Customer Match: targeting with first-party data
Customer Match lets you upload contact details that customers have shared with you — typically email addresses, and optionally phone numbers and postal details — so that Google can match them to signed-in users and let you reach them, or exclude them, across Search, Shopping, YouTube, Gmail, Display and other surfaces. Details are hashed before matching, and Google sets policies on how the data may be used.
Why it matters
As browsers restrict third-party cookies and more visitors decline advertising cookies, website-based lists capture a smaller share of real visitors. Customer Match uses data you already hold with your customers’ knowledge, which makes it more durable. It is particularly useful for existing customers, lapsed customers, leads who have not yet bought, newsletter subscribers and event attendees.
Practical uses
- Excluding existing customers from acquisition campaigns, so new-customer budget is not spent on them.
- Win-back campaigns for customers who have not bought for a defined period.
- Cross-sell and renewal messages to customers of one product or service.
- Lead nurturing for enquiries that did not convert, with messages tailored to their stage.
- Audience signals for automated campaigns, such as Performance Max, to help them find people similar to your best customers.
Getting the basics right
Customer Match is subject to eligibility requirements and policies set by Google, which change over time, so check current terms for your account. Match rates vary; business email addresses often match less well than personal ones, so B2B lists may produce smaller audiences than their raw size suggests. Lists need refreshing regularly, ideally through an automated connection with your CRM rather than occasional manual uploads.
Data protection
Using customer data for advertising is processing personal data under UK data protection law. You need a lawful basis, your privacy notice should explain that you use customer details for advertising on third-party platforms, and individuals who have objected to marketing or asked not to be contacted should be removed. Which lawful basis applies, and whether consent is needed, depends on how the data was collected and the relationship involved, so take advice if unsure and check current guidance from the Information Commissioner’s Office.
Remarketing frequency capping
Frequency capping limits how often the same person sees your ads within a period. Without it, remarketing campaigns tend to concentrate impressions on a small number of heavily exposed people, because they browse widely across sites showing display ads. Beyond a point, extra impressions add cost without adding conversions, and persistent repetition can create a negative impression of the brand.
Where caps can be set
Frequency controls are available for some campaign types, such as display and video, and limited or managed automatically in others, such as Performance Max, where Google’s systems decide exposure. Check which controls apply to each campaign type you use, and where manual caps are not available, use other levers such as audience size, budget and creative rotation.
Choosing a cap
There is no universal right number. A reasonable starting point for display remarketing is a few impressions per person per day, with a weekly limit as well, then adjusting based on performance. High-intent, short-window segments can tolerate more frequent exposure for a few days; large, low-intent segments need tighter limits. Look at performance by frequency where the reporting allows, and reduce caps where additional impressions stop producing results.
Refreshing creative
Fatigue is about the message as well as the count. Rotating several creative variations, refreshing them every few weeks and matching messages to segments all reduce the sense of being followed by the same advert.
Watching for brand damage
Signs of over-exposure include falling click-through rates over time within a segment, rising cost per conversion as frequency climbs, and, occasionally, direct complaints. Placement exclusions also matter: showing ads on low-quality or inappropriate sites can harm a brand regardless of frequency, so review where ads appear and exclude unsuitable placements.
Writing remarketing ads that match each segment
Segmentation only pays off if each segment hears something different. The message should reflect what the person was doing and what is most likely to be holding them back.
For abandoners
Remind them specifically of what they left: the product, the quote, the booking. Address the most common reasons people hesitate at that point, such as delivery times, total cost, installation, warranty or returns. Where you genuinely have one, a time-limited reason to act can work, but avoid inventing urgency; people notice, and it erodes trust. Make the route back as short as possible, ideally to the exact page or basket they left.
For product or service viewers
Relate the ad to the category or service they viewed. Social proof works well here: reviews, ratings, case studies from similar customers, accreditations. Showing the breadth of the range or the outcomes customers have achieved helps move people from interest towards a shortlist.
For low-intent readers
Offer something useful rather than a hard sell: a guide, a checklist, a webinar, a newsletter. The aim is to stay in mind and perhaps move them to a medium-intent action, not to push for a purchase they are not ready to make.
For existing customers
Speak to them as customers. Complementary products, renewals, upgrades, referral schemes and service reminders all make sense; acquisition offers for new customers usually do not, and can annoy people who paid full price last month.
Creative formats and testing
Provide a range of image sizes, headlines and descriptions so ads can appear in more placements, and test variations within each segment. Responsive display ads assemble combinations automatically, which helps, but the underlying messages still need to be written for the segment. Refresh creative every few weeks for long-running segments to reduce fatigue.
Rebuilding remarketing in eight weeks
The timeline sets out a practical sequence for restructuring an account that currently runs a single remarketing list, from consent and tracking through to testing whether the new setup adds conversions.
Consent and exclusions come first because they cost nothing, reduce risk and improve results immediately. Segment lists then need time to fill before campaigns can use them, which is why preparation of creative and Customer Match runs in parallel. The final step, testing incrementality, is the one most often skipped and the one that tells you whether remarketing is genuinely earning its budget.
Remarketing practices in UK accounts
The progress bars show how often each practice is in place across the UK Google Ads accounts we audit. Figures are indicative.
Remarketing practices in place
The pattern mirrors the maturity score. The technical foundation, a tag on every page, is almost universal. Each step that requires a decision — excluding converters, splitting by intent, capping frequency, writing specific messages, using first-party data, testing whether it works — is less common than the one before. Most of the improvement available in a typical account lies in the middle of this list, and none of it requires additional budget.
Dynamic remarketing feeds for e-commerce
Dynamic remarketing shows people ads featuring the specific products or services they viewed, along with related items, pulled automatically from a feed. For online retailers it is often the strongest-performing form of remarketing, because the ad reflects exactly what someone was considering.
How it works
For retail, the product feed in Google Merchant Center supplies product details, images, prices and availability. The Google tag on your site records which products each visitor views, adds to basket or buys, using product identifiers that match the feed. Ads are then assembled automatically for each person from the products they engaged with. Other sectors, such as travel, property, education and jobs, can use dynamic remarketing with a business data feed tailored to their vertical.
Getting the feed right
The dynamic remarketing feed is only as good as its data. Product identifiers in the tag must match the feed exactly, or products cannot be shown. Images should be clear and meet Google’s requirements; prices and availability must be accurate and kept current, because ads showing out-of-stock items or wrong prices waste money and frustrate customers. Titles should be clear and descriptive. Our guide to Google Shopping and remarketing campaigns covers feed optimisation in more depth.
Segmenting dynamic campaigns
The same intent principles apply. Basket abandoners deserve different treatment from product viewers, and recent purchasers should usually be excluded or shown complementary products rather than the item they just bought. Excluding products with low margins or that rarely sell through remarketing can also improve results.
Dynamic remarketing in automated campaigns
Performance Max campaigns with a product feed include dynamic remarketing within their automated approach, using your audience lists as signals. This simplifies setup but reduces control over how much budget goes to remarketing versus new customers. Reporting and, where available, settings for new-customer acquisition help you understand and influence the balance.
Audience signals in automated campaigns
Google’s automated campaign types, particularly Performance Max and Demand Gen, change how remarketing lists are used. Instead of strictly limiting ads to a list, these campaigns often use your lists as signals: hints about the kind of people likely to convert, which the system uses as a starting point before expanding to others it predicts will perform.
This has practical consequences. Your remarketing lists and Customer Match lists become valuable inputs even for campaigns aimed at new customers, because they show the system what good customers look like. At the same time, it can be harder to see how much of an automated campaign’s budget goes to people who already know you versus genuinely new audiences, and remarketing conversions can make an automated campaign look stronger than its new-customer performance justifies.
A few practices help. Provide strong signals: high-intent lists and customer lists rather than broad all-visitor lists. Where available, use settings and reporting that distinguish new from returning customers, and consider valuing new-customer conversions differently. Keep dedicated remarketing campaigns where you want explicit control over segments, messages and frequency. And avoid running several campaigns that all target the same high-intent users without a clear plan, as they may compete with each other and make results hard to interpret.
UK consent requirements: PECR and remarketing tags
This section is a general overview, not legal advice. Check current guidance from the Information Commissioner’s Office and take advice for your circumstances.
Remarketing depends on storing or reading identifiers on visitors’ devices, typically through cookies or similar technologies set by the Google tag. In the UK, the Privacy and Electronic Communications Regulations (PECR) require consent before storing or accessing information on a user’s device, unless it is strictly necessary for a service the user has requested or falls within a specific exemption. Advertising and remarketing cookies are generally not strictly necessary, and the ICO’s guidance treats them as requiring consent. Recent legislative changes have introduced some exemptions for certain low-risk purposes, but advertising is not generally among them.
In practice, this means:
- Remarketing tags should not set advertising cookies before consent. A consent banner that displays a notice while tags fire anyway does not meet the requirement.
- Consent must be freely given, specific and informed. Rejecting advertising cookies should be as easy as accepting them, and pre-ticked boxes or treating continued browsing as consent are not valid.
- The choice must be respected technically. Tag management should check consent status before advertising tags run.
- Records and transparency. Keep records of consent, and explain in your cookie and privacy information what is used, why and by whom.
Google’s consent mode
Google also requires advertisers using personalised advertising features, including remarketing, with users in the UK and European Economic Area to obtain consent and pass consent signals to Google, typically through consent mode. Consent mode lets the Google tag adjust its behaviour according to each visitor’s choices and supports modelled conversion reporting where consent is not given. Without correct consent signals, remarketing lists and measurement for UK users may be limited. Check Google’s current policy and requirements for your account.
Why it matters commercially
Beyond compliance, consent changes the size of your lists. If a significant share of visitors decline advertising cookies, website remarketing lists will be smaller than total traffic suggests, which is one reason Customer Match and well-designed consent experiences matter. Enforcement powers under PECR have also been strengthened, with maximum penalties brought into line with UK data protection law, and the ICO has shown interest in cookie compliance on widely used websites.
How often remarketing tags respect consent
Across the UK business websites we check, the proportion where advertising or remarketing tags fire before the visitor has given consent remains significant.
The figure is indicative rather than a formal survey. The cause is rarely deliberate. Common reasons include a consent banner added without changing how tags are loaded, tags hard-coded into the site template outside the tag manager, agency or plugin installations that bypass consent settings, and consent mode configured with default settings that do not match the banner. Each is fixable, usually in a few hours, and checking is straightforward: load the site in a private browser window, decline cookies, and see which cookies and requests appear.
Remarketing for lead generation and B2B
Much remarketing advice is written for online shops, but lead generation and B2B businesses often benefit as much or more, because their buyers take longer to decide and visit several times before making contact. The principles are the same; the details differ.
- Define intent by content, not baskets. High-intent signals include viewing pricing or service detail pages, case studies, contact or booking pages, and starting but not completing a form. Downloading a guide or attending a webinar can mark someone as medium intent.
- Use longer windows. B2B sales cycles often run for months. Medium-intent durations of 90 days or more may be appropriate, with high-intent lists refreshed more often.
- Lead with evidence. Case studies, client logos where permitted, accreditations and practical guides tend to outperform generic offers for considered purchases.
- Connect the CRM. Customer Match lists of open opportunities, lost deals and existing clients allow messages for each stage, and exclusion of current clients from acquisition campaigns.
- Remember the buying group. Several people at the same organisation may visit; content suited to different roles, such as finance, technical and operational, can help.
- Measure against pipeline, not just form fills. Where possible, import offline conversion data so you can see which segments produce qualified leads and revenue, not just enquiries.
For lower-traffic B2B sites, list sizes are the main constraint. Start with two or three broader segments, use longer durations to build volume, and rely more heavily on search remarketing and Customer Match, where smaller audiences can still be useful.
Case study: from one list to five
A UK home improvement company with around 60 staff, selling and installing garden rooms, spent about £3,500 a month on display remarketing from a single list of all visitors over 30 days. The campaign reported a reasonable number of conversions, but cost per enquiry was more than double that of its search campaigns, and the managing director had begun to receive comments from friends about seeing the same garden room advert “everywhere”.
An audit found that more than half of remarketing impressions went to visitors who had viewed a single page, often a blog post about garden room planning rules that attracted a broad national audience. Existing customers, who visited regularly to check installation dates, were included, as were job applicants. There was no frequency cap, and some visitors were receiving well over twenty impressions a day. Remarketing tags were also firing before visitors made a consent choice, because the tags were hard-coded into the site template.
The company fixed tag loading so advertising tags respected consent, and implemented consent mode. It then replaced the single list with five: brochure request and quote form abandoners (fourteen days), visitors who used the online price configurator (sixty days, reflecting a long decision cycle), product range viewers (sixty days), blog readers (thirty days, with a low budget and a planning guide as the offer), and a Customer Match list of customers and recent enquirers for exclusion. Recent converters, careers visitors and visits under ten seconds were excluded. Each segment received its own creative, and frequency was capped by segment.
Over the following three months, with the same total budget, remarketing cost per enquiry fell substantially and came close to search campaign levels. A holdout test then showed that a meaningful share of conversions previously credited to remarketing among configurator users would have happened anyway, so budget was moved from that segment to quote form abandoners, where remarketing appeared to add more. Website remarketing lists were smaller after the consent fix, which made the Customer Match list more important.
“We thought remarketing was working because it reported conversions. It turned out most of the money went to people who were never going to buy, and some of it went to our own customers. Splitting the lists and capping frequency made the same budget go much further, and nobody tells me they are sick of our adverts any more.” — Managing Director, UK garden room company
The lesson is not that remarketing was the wrong channel. It was that a default setup hid strong and weak segments inside one average, and without consent, frequency and incrementality checks, nobody could see what the budget was really doing.
The 12-point remarketing checklist
The checklist covers the essentials for remarketing that converts without wasting budget or creating compliance risk.
- Check consent first. Advertising tags should not set cookies before consent, and consent mode should pass signals correctly.
- Confirm conversion tracking works. Without it, you cannot judge any segment.
- Exclude recent converters. With a duration that reflects your repurchase cycle.
- Exclude non-buyers. Careers, supplier, investor and login page visitors, and very short visits where appropriate.
- Create a high-intent list. Basket or form abandoners and pricing or contact page viewers, with a short duration.
- Create a medium-intent list. Specific product or service page viewers, with a duration matched to the sales cycle.
- Decide what to do with low intent. A low budget with soft messages, or no remarketing at all.
- Separate customers. Exclude them from acquisition and reach them in their own campaigns, ideally via Customer Match.
- Match messages to segments. Specific creative for each list, refreshed every few weeks.
- Cap frequency. Where the campaign type allows, with tighter limits for large, low-intent lists.
- Use dynamic remarketing for e-commerce. With an accurate feed and identifiers matching the tag.
- Test incrementality. Check how many remarketing conversions would have happened anyway, and rebalance.
Segments and durations are not set once and forgotten. Site changes can break list rules, new pages may need assigning to intent levels, and seasonal patterns change how long intent lasts. A short quarterly review of list sizes, rules, exclusions and performance by segment keeps remarketing working as the site and business evolve.
Common remarketing mistakes
The mistakes below account for most of the waste and risk we find in UK remarketing accounts.
- One all-visitors list. Spend follows volume, which means low-intent visitors.
- No converter exclusion. Paying to advertise to people who have just bought.
- Including customers, staff and job seekers. Easily excluded, often forgotten.
- Default durations everywhere. Thirty days is rarely right for every segment.
- Uncapped frequency. Heavy exposure for a few people, fatigue and brand irritation.
- The same ad for every segment. Too weak for high intent, too pushy for low intent.
- Tags firing before consent. A compliance risk that is often invisible to the business.
- Feed and tag mismatches. Dynamic ads that cannot show products, or show the wrong prices.
- Lists too small to serve. Over-slicing on a low-traffic site leaves lists below minimum sizes.
- Trusting reported conversions without testing. Remarketing often takes credit for conversions that would have happened anyway.
Remarketing targets people who have already shown interest, many of whom would return and buy anyway. Default attribution can credit remarketing for those conversions, making it look more effective than it is. Before increasing remarketing budgets on the strength of reported results, test incrementality through holdouts or experiments, and compare against how often visitors return and convert without remarketing.
Remarketing in search, and measuring what it really adds
Remarketing is often thought of as display banners, but some of its most effective uses are in search. Applying remarketing lists to search campaigns lets you recognise past visitors when they search again. In observation mode, you can see how they perform compared with new searchers and adjust bids accordingly. In targeting mode, you can bid on broader terms only for people who already know you, where the same terms would be too expensive for a general audience. Tailored ad copy for returning high-intent visitors — acknowledging they have looked before, addressing likely objections — can also lift results.
Measuring remarketing properly means going beyond the conversions reported in the account. Compare cost per conversion by segment, not just overall. Look at the time from last ad interaction to conversion, and whether conversions credited to remarketing are clustered among people who were already very close to buying. Where traffic allows, use experiments or holdout groups to measure incremental conversions — the ones that would not have happened without remarketing. Our guide to using remarketing to win back lost visitors covers creative approaches, and attribution more broadly is a topic worth understanding before drawing firm conclusions from any single report.
Setting remarketing budgets and bids
Once segments exist, the question becomes how much to spend on each. There is no fixed proportion of a Google Ads budget that should go to remarketing; it depends on traffic volume, sales cycle and how much remarketing genuinely adds. A few principles help.
Budget follows intent, capped by audience size. High-intent segments justify the highest spend per person, but they are small, so there is a natural ceiling on how much they can usefully absorb. Pushing more budget into a small list mainly raises frequency. Once high-intent segments are saturated, additional budget is usually better spent on acquisition than on low-intent remarketing.
Separate budgets give control. Running segments in separate campaigns, or at least with separate budgets, stops a large low-intent list from consuming spend meant for high-intent visitors. It also makes performance by segment easy to read.
Choose bidding to match the goal. Automated bidding towards conversions or conversion value usually works well for remarketing once conversion tracking is reliable and volumes are sufficient. For small segments with few conversions, simpler strategies or manual limits may be more stable until data builds up.
Set targets per segment. A high-intent segment may comfortably achieve a lower cost per conversion than search, while a low-intent segment may never reach it. Judge each segment against what it can realistically deliver and against its incremental contribution, and be willing to switch off segments that do not earn their place. The same logic applies to budget allocation across the whole account: money should follow what each campaign genuinely adds.
At a glance: Google Ads remarketing
| Area | Recommended approach | Why it matters |
|---|---|---|
| Consent | Ad tags blocked until consent; consent mode implemented | PECR compliance and Google policy requirements |
| Conversion tracking | Verified and accurate | Needed to judge any segment |
| Exclusions | Recent converters, non-buyers, very short visits | Removes a large share of waste immediately |
| High-intent segment | Abandoners and pricing or contact viewers, short duration | Most likely to convert; deserves most budget |
| Medium-intent segment | Product or service viewers, duration matched to sales cycle | Interest in something specific |
| Low-intent segment | Low budget, soft message, or excluded | Large lists where waste builds fastest |
| Customers | Excluded from acquisition; own campaigns via Customer Match | Different goals: retention, cross-sell, renewal |
| Customer Match | CRM-connected lists with lawful basis and privacy notice | Durable first-party audiences as cookies decline |
| Messages | Specific creative per segment, refreshed regularly | Relevance and reduced fatigue |
| Frequency | Capped where available, tighter for large lists | Controls cost and protects the brand |
| Dynamic remarketing | Accurate feed, matching identifiers, segmented by stage | Strongest format for e-commerce |
| Search remarketing | Lists on search in observation or targeting mode | Recognises returning searchers |
| Measurement | Performance by segment and incrementality testing | Remarketing over-claims without testing |
How Cloudswitched approaches remarketing
We start every remarketing engagement with the foundations: checking that advertising tags respect consent, that consent mode is set up correctly and that conversion tracking is accurate. Only then do we look at audiences, because segments built on unreliable data or non-compliant tags create problems later.
We then restructure remarketing around intent and funnel stage: exclusions for converters and non-buyers, separate high, medium and low-intent lists with durations matched to your sales cycle, Customer Match connected to your CRM where eligible, specific creative for each segment, frequency controls appropriate to each campaign type and, for retailers, dynamic remarketing built on an accurate feed. We apply remarketing lists to search as well as display and video, and we report performance by segment so budget can move to where it works.
Finally, we test whether remarketing is genuinely adding conversions rather than taking credit for them, using experiments and holdouts where volumes allow. No advertising setup can guarantee results, but remarketing designed around intent, consent and honest measurement usually makes the same budget work considerably harder.
Want remarketing that earns its budget?
Talk to us about a remarketing audit, audience segmentation, Customer Match or a consent review for your Google Ads account.
Talk to a Google Ads SpecialistFrequently Asked Questions
What is Google Ads remarketing?
A way to show ads to people who have previously interacted with your business — visited your website, used your app, watched your videos or shared their contact details — using audience lists in Google Ads campaigns across search, display, video, shopping and other formats.
Why is a single remarketing list a problem?
Because low-intent visitors usually make up most of the list, they receive most of the impressions and spend, while high-intent visitors get the same generic message. Segmenting by intent lets budget and messages follow likelihood to convert.
How should I segment remarketing audiences?
By what visitors did: high intent (basket or form abandoners, pricing and contact viewers), medium intent (product or service page viewers), low intent (blog readers and brief visits), and existing customers, with exclusions for recent converters and non-buyers.
How long should remarketing lists last?
Match membership duration to how long the intent behind the list lasts. High-intent lists often use short durations, from a few days to a couple of weeks for e-commerce; medium-intent lists can run longer, and businesses with long sales cycles may need longer windows.
What is Customer Match?
A Google Ads feature that lets you upload hashed customer contact details, such as email addresses, to reach or exclude those customers when they are signed in to Google services. Eligibility requirements and policies apply, and you need a lawful basis under data protection law.
What frequency cap should I use for remarketing?
There is no universal number. A few impressions per person per day with a weekly limit is a common starting point for display, adjusted by segment and performance. Some campaign types manage frequency automatically rather than allowing manual caps.
What is dynamic remarketing?
Remarketing that shows people the specific products or services they viewed, pulled automatically from a feed such as a Merchant Center product feed, matched to the products recorded by the Google tag on your site.
Do I need consent for remarketing in the UK?
Generally yes. PECR requires consent before storing or accessing information on users’ devices unless an exemption applies, and advertising and remarketing cookies are generally treated as requiring consent. Google also requires consent signals for personalised advertising with UK users. Check current ICO guidance and take advice for your circumstances.
What is Google consent mode?
A way for your consent banner to pass visitors’ choices to Google tags, so that tags adjust their behaviour accordingly. It supports compliance with Google’s consent requirements and enables modelled reporting where consent is not given.
Why are my remarketing lists smaller than my traffic?
Common reasons include visitors declining advertising cookies, browser restrictions on tracking, list rules that exclude many visitors, and recent list creation. Customer Match can help reach customers who are not captured by website lists.
Does remarketing work for B2B businesses?
Yes, often well, because B2B decisions take time and involve several visits. Longer durations, segments based on service pages and case studies viewed, and Customer Match for leads and contacts tend to suit B2B, though business email addresses may match less well.
How do I know if remarketing is actually working?
Look at performance by segment rather than overall, and test incrementality with experiments or holdout groups to see how many conversions would have happened without remarketing. Reported conversions alone tend to overstate its effect.
Related reading
More guidance for UK businesses from the Cloudswitched blog:
- Link building for UK businesses: earning backlinks without penalties
- Website security hardening: protecting your site from common attacks
- IT governance for UK SMEs: board reporting and risk registers
- Network change management: making changes without outages
- Choosing an office location: evaluating IT infrastructure before you sign
Stop showing the same ad to everyone
Cloudswitched builds Google Ads remarketing around intent, consent and honest measurement, so your budget reaches the people most likely to come back and buy.
Talk to a Google Ads Specialist
Cloudswitched
London-based managed IT services provider offering support, cloud solutions and cybersecurity for SMEs.
