Remarketing in Google Ads: Win Back Lost Visitors
Most of the people who visit your site leave without doing a thing. In B2B that is normal, not a failure. A first visit rarely ends in a demo request when the buying committee has three people, a budget cycle, and a procurement process to clear. The deal takes weeks or months.
Remarketing is how you stay in front of those visitors instead of paying to acquire them all over again. Done well, it is one of the cheapest sources of qualified pipeline you have, because you are talking to people who already raised their hand by showing up.
Done badly, it burns money following the wrong people with the wrong message until they start to resent your brand. This guide covers the difference: how to build audiences that mean something, what to say at each stage, how often to show up, and how to know whether any of it actually paid off.
What remarketing is, and where it runs
Remarketing shows ads to people who already interacted with your business: visited a page, watched a video, used your app, or sit in a customer list you uploaded. Google matches those people across its inventory and serves them your ads.
You have a few formats to work with, and they behave differently:
- Display remarketing. Banner and responsive ads across the Google Display Network, millions of sites and apps. Cheap, broad reach, easy to overdo.
- Remarketing Lists for Search Ads (RLSA). Not a separate ad. You attach a remarketing audience to your search campaigns so you can bid higher, or show different copy, when a past visitor searches again. This is quietly one of the most effective B2B tactics because it catches active intent.
- Video remarketing on YouTube. Re-engage people who watched your videos or visited your site, with skippable in-stream or short formats.
- Performance Max. Includes remarketing-style audience signals, though you give up granular control over who sees what.
If you are still deciding how display fits against search overall, the trade-offs in Google Search vs Display Network are worth reading before you pour budget into banners.
A naming note. Google's own UI now leans on "audiences" and "Your data segments" rather than the old "remarketing list" language. The concept is identical. Older tutorials calling it "remarketing lists" are describing the same thing.
Why B2B remarketing is different
Consumer remarketing chases a quick repurchase: you looked at running shoes, here are the running shoes again, buy now. B2B does not work like that.
Your visitor is one person on a committee. The decision is slow and rational. Nobody signs a contract because they saw a banner three times. So the job of B2B remarketing is not to close. It is to stay credible and present long enough that when the buying window opens, you are the name they already trust.
That changes everything about how you segment and what you say. A blog reader who spent ninety seconds on one article is not the same prospect as someone who viewed your pricing page twice and started a demo form. Treating them identically is the most common mistake, and it is the one that wastes the most money.
Build audiences around intent, not just "all visitors"
The default Google offers, all visitors over the last 30 days, is the worst possible segment to spend on. It mixes a procurement lead with a job applicant with someone who clicked a link by accident.
Segment by what the visit signals about buying readiness. A practical structure:
| Segment | Signal | Message | Window |
|---|---|---|---|
| Blog and top-of-funnel readers | Read one article, low intent | More useful content, a guide, light brand | 14 to 30 days |
| Solution and service pages | Looked at what you do | Case study, proof, a clear next step | 30 to 60 days |
| Pricing or contact page | Strong commercial intent | Direct offer: book a call, get an audit | 30 to 90 days |
| Form abandoners | Started, did not finish | Remove friction, reassure, one-click return | 7 to 14 days |
| Existing customers | Already bought | Exclude from acquisition, target for upsell only | Ongoing |
The membership windows matter. A 540-day maximum is available, but a B2B prospect who looked at pricing six months ago and went silent is usually cold. Long windows make sense for high-consideration purchases with long cycles; short windows keep your spend on people whose interest is still warm. Match the window to your real sales cycle, not to the maximum Google allows.
One segment deserves special attention: form abandoners. Someone who started a demo request and bailed is the closest thing to a buying signal you get from a website. A tight, urgent, short-window campaign aimed only at them often returns the best cost per lead in the whole account.
Set up the data source first
None of this works without clean tracking. You need the Google tag (gtag.js) or Google Tag Manager firing across the site, and your conversion events defined correctly. If your event setup is shaky, your audiences will be shaky too, and you will optimize toward noise.
Build audiences from URL rules (anyone who hit /pricing), from events (started a form, scrolled past 75%), or from time on page. Customer Match, uploading a list of emails from your CRM, lets you remarket to known contacts, though it needs a reasonable list size to activate.
Match the message to the stage
The fastest way to waste remarketing budget is to show everyone the same "Book a demo" banner. A blog reader is not ready for that ask. A pricing-page visitor is annoyed by anything softer.
Think of it as a short sequence rather than one repeated shout:
- Early visitors get value and credibility. A useful resource, a customer logo wall, a short proof point. You are earning the right to ask, not asking yet.
- Mid-funnel visitors get evidence. A case study with a number, a comparison, an answer to the objection you know they have.
- High-intent visitors get the direct offer. Short, specific, low friction. "Get a free 20-minute audit," not "Learn more."
This is also where RLSA earns its keep. When a past pricing-page visitor types a commercial search query, layering a remarketing audience onto that search campaign lets you bid up and tailor the headline to someone who already knows you. The intent is fresh and the cost to convert is usually far lower than a cold click. If you want the broader picture of running search for this audience, the complete Google Ads for B2B guide ties the channels together.
Frequency, exclusions, and not being creepy
Show your ad too often and you cross from "present" to "stalking." There is no universal magic number, but a frequency cap in the range of a few impressions per user per day, or per week for display, keeps you visible without fatiguing the audience. Watch the frequency column in your reports and pull back when click-through rate falls while impressions climb.
Exclusions matter as much as targeting. Always exclude:
- People who already converted, unless the campaign is built for upsell.
- Existing customers from acquisition campaigns.
- Job seekers and your own team, if you can identify the pages they visit.
- Placements that send junk. Mobile game apps are a classic budget drain for B2B; review your placement report and exclude aggressively.
Privacy rules are not optional. With third-party cookies degrading and consent requirements tightening across the US and Europe, your audiences depend on a working consent setup and first-party data. Make sure your consent banner and Google's consent mode are configured, or your lists will quietly shrink and your measurement will drift.
Measure by pipeline, not by view-through
Here is where most remarketing reports lie to themselves. Display remarketing loves to claim credit through "view-through conversions," counting a conversion because someone saw a banner they never clicked, then converted later. Some of that influence is real. A lot of it is the campaign taking credit for demand it did not create.
Judge remarketing the way you judge every other channel: by qualified leads and pipeline, not impressions or assisted soft metrics. Connect your conversions to real outcomes downstream. The honest questions are simple. Did remarketing-touched leads turn into opportunities? At what cost? Would those people have come back anyway?
That last question is the hard one, and the only clean answer is testing. Hold out a portion of an audience, show them nothing, and compare conversion rates against the treated group. If the treated group converts meaningfully better, your remarketing is doing work. If not, you are paying to reach people who were coming back regardless. Run that test before you scale spend.
Visited site ──► Left without converting ──► Remarketing audience
│
┌─────────────────────────────┼─────────────────────────────┐
▼ ▼ ▼
Display + Video RLSA Form abandoners
(stay present) (catch active search) (close the gap)
└──────────────┬───────────────┴──────────────┬───────────────┘
▼ ▼
Returns and converts Measured against a holdout
Set your bidding to follow the same logic. If you have solid conversion data flowing in, automated strategies like target CPA can do real work here, since the audience is warm and the signals are strong. The when and how of letting the machine bid is covered in the Smart Bidding guide.
Common mistakes that drain the budget
A few patterns show up in almost every underperforming remarketing account:
- One giant "all visitors" audience. No segmentation, one message, wasted spend on people who will never buy.
- No exclusions. Paying to advertise to customers who already signed, or to the same handful of people 40 times.
- Windows that never close. Chasing visitors from eight months ago who have long forgotten you.
- Judging success by impressions or view-throughs. Big numbers, no pipeline.
- Ignoring placements. Letting Google spray banners across low-quality apps and never checking the report.
- Set and forget. Remarketing audiences decay. Creative fatigues. A campaign you built last quarter and never touched is probably leaking money now.
FAQ
What is the difference between remarketing and retargeting?
Nothing meaningful. Google calls it remarketing, much of the industry says retargeting, and they describe the same thing: showing ads to people who already interacted with you.
How long should someone stay in a remarketing audience?
Match the window to your sales cycle. For high-intent segments like pricing-page visitors, 30 to 90 days is common. Form abandoners want a much shorter, urgent window of a week or two. The 540-day maximum is rarely the right choice for B2B, because interest from that long ago is usually cold.
How big does my audience need to be?
Display remarketing typically needs at least 100 active users in the last 30 days to start serving, and RLSA needs around 1,000. Customer Match needs a sizable uploaded list to match against. Smaller B2B sites often struggle to hit these thresholds on narrow segments, so you may need to combine segments or widen windows until traffic grows. Check Google's current minimums, since they change.
Is remarketing worth it for a low-traffic B2B site?
Sometimes not, at least not yet. If you only get a few hundred visitors a month, your audiences may never reach serving size, and your budget is better spent building that traffic first through search and content. Once you have steady volume, remarketing becomes one of your most efficient channels.
Does remarketing still work without third-party cookies?
It works, but it leans harder on first-party data and consent. Customer Match (your own email lists) and on-site audiences built through the Google tag with proper consent are more durable than third-party-cookie tracking. Get your consent mode and first-party data house in order, and remarketing keeps performing.
How do I keep remarketing from annoying people?
Cap frequency, exclude people who already converted, keep windows tight, and refresh creative before it goes stale. If your click-through rate is dropping while impressions rise, you are showing the same ad too often. Pull back.
Bringing it together
Remarketing rewards precision. Segment by intent, write a message that fits each stage, exclude the people you should not pay to reach, cap how often you appear, and measure against real pipeline with a holdout to prove it. Skip those steps and you get a banner habit that quietly drains budget while looking busy in the dashboard.
A quick checklist before you launch:
- Conversion tracking and consent mode working, audiences built from intent signals
- Separate segments for readers, solution-page visitors, pricing visitors, and form abandoners
- Message matched to each stage; RLSA layered onto search for active intent
- Exclusions in place: converters, customers, junk placements
- Frequency capped, windows matched to your real sales cycle
- Success measured by qualified leads and pipeline, validated against a holdout
If your remarketing is running on autopilot and you are not sure it is paying off, that is worth fixing before you spend another month on it. We can audit your Google Ads account, show you exactly where the budget leaks, and lay out which audiences are earning their keep. Reach out for a focused review and you will leave knowing what to cut and what to scale.