Remarketing
Google's name for retargeting. Same idea: ads shown to people already on one of your lists.
Retargeting ads (Google calls them remarketing) are shown only to people who already visited your website or are on your customer list. They usually look like your cheapest campaign because the platform also takes credit for customers who were coming back anyway: in a published field experiment, display retargeting lifted purchases by 11%, or about one sale in ten among the people who saw it. A small business gets real value from retargeting by excluding past buyers and job seekers, matching the list length to its buying cycle, capping frequency, and judging it on new paying customers.
Free account review
Tell us where you are. We reply within two business days with what we would fix first.
01 — The short answer
Retargeting ads are ads that only reach people who have already interacted with your business, usually by visiting your website. A small piece of code on your site adds each visitor to a list, and Google or Meta then shows your ads to that list as those people browse, search or scroll. In the U.S., "retargeting" gets about 1,600 Google searches a month, and "retargeting ads", "Google retargeting ads" and "Google Ads remarketing" about 880 each (Google Keyword Planner, October 2026).
Retargeting is not a separate platform but an audience you attach to campaigns on Google (Display, YouTube, Search, Demand Gen, Performance Max) and Meta (Facebook and Instagram). The useful question for an owner is how many of the customers it reports you would have lost without it.
Google's name for retargeting. Same idea: ads shown to people already on one of your lists.
A small piece of code on your website that tells the ad platform who visited and which pages they saw.
How many days a visitor stays on a list after their last visit before dropping off.
A sale or lead credited to an ad the person saw but never clicked.
A list you tell the platform never to show a campaign to, such as people who already bought.
The share of results that would not have happened without the ad. It is the only share worth paying for.
02 — Google vs Meta
Both platforms build lists from your website tag and from customer lists you upload, but the limits differ. A Google remarketing list needs at least 100 active users in the last 30 days before it can serve on Search, Display or YouTube, keeps people for 30 days by default and up to 540 days, while a Meta website custom audience keeps people for 1 to 180 days (Google Ads Help and Meta Marketing API documentation, checked October 2026).
| Rule | Google Ads | Meta (Facebook and Instagram) |
|---|---|---|
| What it is called | Remarketing, "your data segments" | Website custom audience |
| Uploading your customer list | Customer Match | Customer list custom audience |
| Minimum list size to serve | 100 active users in the last 30 days (Search, Display, YouTube) | No single published minimum; small audiences deliver slowly |
| How long a visitor stays | 30 days by default, up to 540 | 1 to 180 days |
| Frequency cap | Per day, week or month; on Display only viewable impressions count | Hard cap only on Reach campaigns; otherwise watch the Frequency column |
| Default view-through credit | 1-day window for new conversion actions | 1-day view is one of the attribution settings still offered |
Sources: Google Ads Help (lists, frequency, view-through); Meta custom audiences; checked October 2026.
The 100-user rule matters more than most guides admit: a local business whose site gets a few hundred visitors a month can have Google remarketing switched on and barely serving.
03 — Why it looks cheap
Yes, but much less than the report says. In a field experiment on an online retailer's display retargeting campaign, Johnson, Lewis and Nubbemeyer measured a 17% lift in site visits and an 11% lift in purchases (working paper 2016, Journal of Marketing Research 2017). That means about one in ten purchases from people who saw the ads was caused by them; a standard ad report can credit all ten.
| Group | Purchases | What it means |
|---|---|---|
| People who never saw the ads | 100 | The baseline: customers who came back on their own. |
| People who saw the ads | About 111 | An 11% lift, the published result. |
| Sales the ads caused | About 11 of 111 | Roughly one in ten. The other nine would have happened anyway. |
Illustration of the published 11% lift, scaled to 100 baseline purchases. Source: Johnson, Lewis & Nubbemeyer, "Ghost Ads" (read 19 October 2026).
Three things inflate the report. Existing customers who visit to pay an invoice land on the list and get counted when they buy again. View-through credit counts sales from people who scrolled past a banner without clicking; in Google Ads that window is one day by default. And automated bidding learns from those credited sales, so it spends more on the people who needed you least. We call money judged this way The Blind Budget: ad spend graded by what the platform counts instead of by customers you would not have had without it.
04 — Who to exclude
Exclude anyone who visits your site for a reason other than buying: people who already paid you, job seekers, and customers who only log in or pay a bill. Each one makes the reported cost per sale look better than the truth.
| Exclude | Why | How |
|---|---|---|
| People who already bought | They are the easiest "conversions" for the platform to claim, and they were coming back anyway. | Upload your customer list: Customer Match on Google, a customer list custom audience on Meta. Add recent form submitters by the thank-you page URL. |
| Job seekers | A careers page can be one of the most visited pages on a local service site. | A list of everyone who viewed the careers or jobs URL, set as an exclusion. |
| Login and payment visitors | Customers paying an invoice or booking a repeat visit are not new demand. | Lists of people who only viewed the login, portal or payment pages. |
Before uploading a customer list to Google, check that your privacy policy says you share customer data with third parties such as ad platforms, and get consent where the law requires it; both are conditions of Google's Customer Match policy (checked 21 October 2026). On Performance Max, Google's customer acquisition goal has a "New Customer Only" mode, which Google recommends for lead generation, using your customer list and website tag to spot existing customers (Google Ads Help, 21 October 2026).
05 — How long
As long as your customers usually take to decide, and no longer. Google's default list duration is 30 days and Meta allows up to 180, but an emergency plumber's customer decides in hours while a kitchen remodel can take months, so one default fits neither. Find your buying cycle in your own sales records: the time from a customer's first visit or call to the day they pay.
| Buying cycle | Example businesses | Starting list duration |
|---|---|---|
| Hours | Emergency plumbing, locksmith, towing | 3–7 days, or skip retargeting |
| Days | AC repair, auto repair, pest control | 14–30 days |
| Weeks | Dental implants, med spa treatments, moving | 30–90 days |
| Months | Roof replacement, kitchen remodel, home care | 90–180 days |
Starting points Eloari uses before an account has its own data, not platform rules. Adjust them to your records.
Browsers shorten every window. Since Apple's Intelligent Tracking Prevention 2.1 update in 2019, Safari limits cookies set by a page's script to seven days (WebKit), so many iPhone visitors fall off a list within a week whatever duration you choose. Chrome is looser: in April 2025 Google said Chrome would keep third-party cookies under people's existing settings.
06 — Frequency and health
Cap how often one person sees your ads: Google Ads lets you set a limit per day, week or month, while most Meta campaigns have no hard cap, so you watch the Frequency column and change the ad when it climbs while sales stay flat. Health businesses face a stricter rule: health is a sensitive category in Google's personalized ads policy, so those advertisers cannot use their own remarketing lists or Customer Match on Google (Google Advertising Policies Help, October 2026).
Frequency is the average number of times one person saw your ad; on a small list the same handful of people can see the same banner every day. If you run a clinic, dental practice or med spa and someone tells you they are "retargeting your past patients on Google", ask how, because Google's personalized advertising policy does not allow advertiser-curated audiences in that category. See healthcare PPC and Google Ads for dentists for what works instead.
07 — The test
Run a holdout test: leave part of your service area out of retargeting on purpose for six to eight weeks and compare new paying customers from your own books. Google's own Conversion Lift studies do the same thing at scale, but they carry spend and conversion minimums that most small accounts do not meet.
| Step | Time | What to do |
|---|---|---|
| Pick two matching areas | Day 1 | Two parts of your service area (ZIP codes or towns) that usually bring about the same number of customers. |
| Exclude one | Day 1 | Keep retargeting on in one area and exclude the other location from every retargeting campaign. |
| Wait | 6–8 weeks | Longer if your buying cycle is long. |
| Count customers, not conversions | End of test | Count new paying customers in each area from your invoices or CRM, not from the ad report. |
| Decide | End of test | About the same? Retargeting was mostly paying for people already coming back. Clearly more in the retargeted area? Now you know what it is worth. |
A rough test: compare the same weeks and keep budgets and offers equal in both areas.
08 — Verdict
Run retargeting when your site has enough visitors to keep lists above Google's 100-user minimum, your buying cycle is longer than a day, and you can exclude existing customers. Skip it or pause it when the list barely serves, when customers decide within hours, or when Google's health rules bar you from your own lists.
| Situation | What it means | Verdict |
|---|---|---|
| Fewer than 100 visitors in 30 days | Google lists cannot serve on Search, Display or YouTube. | Fix traffic first. Spend on search or Meta prospecting. |
| Customers decide within hours | Most retargeted visitors already hired someone. | Short window (3–7 days) or none. |
| High ticket, weeks or months to decide | Visitors are still comparing quotes. | Worth running, with exclusions and a long window. |
| Retargeting is your "best" campaign | It may be claiming returning customers. | Run a holdout before raising its budget. |
09 — Management fee
Retargeting is part of managing your Google Ads and Meta Ads accounts, not an add-on. Eloari charges a flat monthly fee from $900, never a percentage of ad spend, month to month with no setup fee, so cutting retargeting that does not pay never cuts our income.
| Tier | Monthly fee | Suited to |
|---|---|---|
| Starter | $900/mo | Ad spend under $5,000/mo |
| Growth | $1,500/mo | Ad spend $5,000–$20,000/mo |
| Full | $2,500/mo | Ad spend $20,000+/mo |
You pay Google and Meta directly, with no markup, and you own every ad account, pixel, audience list and conversion history. Tracking is installed and verified before any spend begins, so retargeting is judged on real sales and booked jobs, not page visits.
10 — Who runs it
Jonatã Ariel Silveira
President, Eloari Management Group · Meta Business Partner
Accounts are run by the founder, not passed to a junior after the sales call. In a retargeting review, the first screen we open is the audience list: who it includes, who it leaves out, and for how long. Eloari currently manages more than $100,000 per week in advertising spend across client accounts.
12 — Questions
Retargeting ads are ads shown only to people who already visited your website or are on a customer list you uploaded. A tag on your site adds each visitor to a list, and Google or Meta shows your ads to that list. Google calls it remarketing; Meta calls the list a website custom audience.
For an advertiser buying ads, nothing: remarketing is Google's name for retargeting.
They work, but less than ad reports suggest. In a field experiment on an online retailer's display retargeting, the ads lifted purchases by 11% (Johnson, Lewis and Nubbemeyer, 2016), so about one in ten purchases from people who saw them was caused by the ads. Test it with a holdout area before raising the budget.
There is no separate price: retargeting runs inside your Google or Meta campaigns and you pay the same auction prices, usually on a smaller audience. The real cost to watch is cost per new customer the ads caused, which you only learn by comparing an area with retargeting against one without it.
Not with their own lists on Google. Health is a sensitive category in Google's personalized advertising policy, so health advertisers cannot use advertiser-curated audiences such as remarketing lists or Customer Match. Check Meta's own health rules before using any patient data there.
Yes, unless the campaign is meant to sell them something new. Upload your customer list (Customer Match on Google, a customer list custom audience on Meta) as an exclusion; on Performance Max, Google's New Customer Only mode uses that list and your tag to focus on new customers.
13 — Next step
Tell us whether you retarget on Google, Meta or both, and roughly how many new customers you served last month. We will check who your lists include, who they should exclude and how the account gives itself credit, whether or not you hire us.