Google Customer Match for B2B.
Google Customer Match for B2B: the identifiers Google accepts, the $50,000 gate on Targeting, why work-email uploads match badly, and the landmine play.
Google Customer Match lets you upload a contact list to Google Ads and reach those exact people on Search, YouTube, Gmail, Display and the Shopping tab. For B2B it breaks in two places: your CRM holds work emails and Google matches against personal Google accounts, and the Targeting setting is locked behind $50,000 of lifetime account spend.
Everything else people write about Customer Match for B2B is downstream of those two facts. This page covers what Google actually accepts today, what the two access tiers let you do, why the match rate collapses on a raw CRM export, the one play that makes the channel worth the effort, and the parts that will not work for you.
All Google-side facts below were checked against Google's own live documentation on 2026-07-29, with links.
What is Google Customer Match?
Customer Match is Google Ads' first-party-data audience feature. You upload contact information your customers gave you, Google hashes it and compares it against signed-in Google accounts, and the resulting matched audience becomes something you can target, observe or exclude in a campaign.
Google's eligibility page states plainly that it is "currently available on Search, the Shopping tab, YouTube, Gmail, and Display" (Google Ads Help, About Customer Match).
It was built for consumer brands with large customer files. That design decision is the source of every B2B problem on this page.
What data does Google Customer Match accept?
Google's upload template accepts "a combination of the following header names in English: 'Email,' 'Phone,' 'First Name,' 'Last Name,' 'Country,' and 'Zip'" (Google Ads Help, Create a Customer Match list). Mobile device IDs use a separate template and should be unhashed. Through the API you can also send CRM user IDs you assigned yourself (Google Ads API, Customer Match).
| Identifier | Accepted | Hashed with SHA-256 | Notes for B2B |
|---|---|---|---|
| Yes | Yes | The field everyone uses. Work email is the wrong one — see below. | |
| Phone | Yes | Yes | E.164 format with country code. Mobile numbers match far better than switchboards. |
| First name + last name | Yes, but only as part of an address | Names yes, country and zip no | Google requires all four address columns together: "For mail address matching, your data file must contain these 4 column headers: 'First Name,' 'Last Name,' 'Country,' and 'Zip'" (Google Ads Help). A name on its own is not a matchable identifier. |
| Country | Yes | No | ISO two- or three-letter code. |
| Zip / postal code | Yes | No | US and international both accepted. |
| Mobile device ID (IDFA / AAID) | Yes | No — upload unhashed | Separate template. Useful where you hold app or ad-platform IDs. |
| CRM / user ID | Yes, via API | n/a | Only matches users you have already tied to a Google identity. |
Two details that trip people up. Country and zip are not hashed even in an otherwise-hashed file. And Google says the matching process "can take up to 48 hours" (Google Ads Help, About the customer matching process), so an audience that looks empty an hour after upload is not necessarily a failed upload.
Why do B2B Customer Match rates come out so low?
Because of one sentence: Google matches against Google accounts, and Google accounts are personal.
Your CRM has sarah.chen@acme.com. Sarah's Google account is sarahchen98@gmail.com, made in 2011, tied to her phone, her YouTube history and her Android device. Google is not withholding the match. It has no bridge between those two strings.
This is the same structural failure that wrecks Meta uploads, and it is the reason B2B Facebook ads underperform for most teams. Work identity and consumer identity are two different graphs, and every consumer ad platform indexes the second one.
Uploading a raw work-email export to Google Ads typically matches around 30% of a B2B list, in what we see across ContactLevel accounts.
That is our own number, from our own accounts, and it is higher than the published third-party figures. Primer reports that uploading a work-email list to Google "might yield a mere 2% match rate" (Primer), and Store Growers reports 29-62% on typical mixed lists (Store Growers). Our B2B match rates page carries those citations. We are not going to pretend the spread away: the lists we see are B2B SaaS lists that have usually had some enrichment history, a mixed consumer list behaves differently again, and nobody in this market publishes their methodology. Treat ~2% as the floor for a raw work-email-only upload and ~30% as what a reasonably clean B2B list does in our accounts. Run a match test on your own list rather than trusting either number.
The fix is not a better export. It is more identifiers per person. Take the one thing you have — the work email, or the LinkedIn URL, or the name plus company domain — and resolve it to the personal identifiers Google indexes: personal email, mobile number, postal address. One CRM row becomes 50-70 data points, and the contact-info upload carries several matchable identifiers for the same human instead of one unmatched one. (Device IDs are a separate upload — a Customer Match list takes one key type, contact info or mobile advertising IDs or CRM IDs, not a mix.) That is what moves ContactLevel-synced lists into the 70-99% range rather than 30%.
The mechanics are the same on every platform and are broken down in contact-level targeting and B2B match rates.
Do you need $50,000 of spend to use Google Customer Match?
This is the single most under-reported constraint in the channel, and it decides whether any of the plays below are available to you.
Google's Customer Match policy defines two tiers (Google Ads Help, Customer Match policy):
| Basic access | Enhanced access | |
|---|---|---|
| Who qualifies | "All policy-compliant advertisers" | "90 days of Google Ads history and more than USD $50,000 total lifetime spend" |
| Observation setting | Yes | Yes |
| Exclusions | Yes | Yes |
| Targeting setting | No | Yes |
| Manual bid adjustments | No | Yes |
Read the middle row again. Without enhanced access you can observe your ICP list and you can exclude it. You cannot restrict a campaign to it.
That matters because almost every "use Customer Match to filter Search to your ICP" article on the internet assumes the Targeting setting is simply a dropdown. It is a dropdown that appears after $50,000. That includes our own Google Ads guide, which does not yet mention the gate — this page is the corrected version.
If you are under the threshold, the two things Customer Match still buys you are real:
→ Exclusions. Upload your existing customers and your closed-lost list and exclude them from every campaign. This works on basic access and it is the cheapest waste-reduction available in a B2B Google account. See Google branded search exclusion.
→ Observation plus reporting. Layer the ICP list as an observation audience and read the segmented performance. You cannot restrict delivery, but you can find out what share of your spend is currently landing on people you would actually sell to. For most B2B accounts the answer is unpleasant and useful.
The landmine play: how do you put your ICP list on Google Search?
Here is the play worth the setup, once you have enhanced access.
You build an audience of the exact people you want — named contacts at target accounts, filtered by role — enrich it so it actually matches, and sync it to Google as a Customer List. Then you attach it to a Search campaign with the audience setting on Targeting, not Observation, and you bid on your category terms.
Now the campaign is a landmine. It sits there costing nothing. The moment one of those named people searches "buyer intent software" or your competitor's name, it detonates and your ad is the result they see. Someone outside the list runs the identical query and you pay nothing.
Why it works for B2B specifically:
→ It inverts the usual Google problem. Normally you buy a keyword and take whoever searches it — students, job seekers, competitors, agencies pitching you. Here you buy a person and wait for the keyword.
→ The economics get better as CPCs rise. At $30-80 CPC in competitive B2B SaaS categories, filtering out the 90-odd percent of searchers who will never buy is worth more than any bid optimisation you can run.
→ It catches the part of the buying committee you never sourced. Your rep is emailing the champion. The landmine catches the VP nobody in sales has spoken to, at the moment she starts researching. That is the buying group problem showing up in a channel where you can actually do something about it.
Two setup details that decide whether it works at all. Set the audience to Targeting, not Observation — on Observation, Google reports on the segment and delivers to everyone anyway, which is how most people conclude the play does not work. And keep the campaign's keyword set tight; the audience filters who sees the ad, it does not fix a bloated broad-match keyword list.
Step-by-step: Google search inclusion and Google Customer Match for B2B.
What is the real minimum list size?
Almost every guide on this topic, ours included, has said Google requires 1,000 matched users before an audience activates.
We could not find that number on any current Google Ads help page as of 2026-07-29. What Google publishes instead:
| Source | What it says |
|---|---|
| About Customer Match | "To stay eligible, a list must have at least 100 members added or updated within the last 540 days." |
| Fix Customer Match issues | "We recommend audience list sizes of at least 100 users to prevent ads from not serving." |
| Google Ads API docs | "Upload at least 5,000 members to increase the chance of having enough matched, active users for targeting." |
Those are three different numbers doing three different jobs: an eligibility floor, a serving recommendation, and a practical recommendation for having enough simultaneously active users. Google is explicit that active users will be fewer than list size, because not everyone is signed in when your ad would serve.
The B2B takeaway is unchanged even though the number moved. A 500-contact target-account list at 30% native match gives you 150 matched records and an audience that will barely serve. The same list at 70-99% gives you 350-495 and a campaign that runs. Enrichment is what gets you over the line, whichever line Google is drawing this quarter.
What changed in Google Customer Match in 2025 and 2026?
2025-04-07 — membership duration. Customer Match lists lost infinite membership duration. Membership spans now run 0 to 540 days (Google Ads API docs). A list you uploaded once and never refreshed decays out from under you. Continuous CRM sync stopped being a nice-to-have on that date.
2026-04-01 — API uploads. Google states: "Starting April 1, 2026, OfflineUserDataJobService and UserDataService requests for Customer Match will fail if the developer token hasn't previously sent requests for Customer Match" (Google Ads API docs). The migration path Google names is the Data Manager API. If your list gets to Google through a tool built after that date on a fresh developer token, that tool needs to be on Data Manager, not the legacy services. Worth one question to any vendor selling you Google sync.
Google Customer Match vs LinkedIn and Meta: what does each one match on?
Same contact list, three platforms, three completely different failure modes.
| Google Customer Match | LinkedIn Matched Audiences | Meta Custom Audiences | |
|---|---|---|---|
| Matches against | Signed-in Google accounts (mostly personal) | LinkedIn accounts (work identity) | Facebook/Instagram accounts (personal) |
| Raw work-email upload | Poor — personal-identity graph | Best of the three — LinkedIn holds work email | Poor — personal-identity graph |
| Native B2B targeting | None | Full: title, seniority, company, industry | None since 2022 |
| Access gate | $50,000 lifetime spend for the Targeting setting | None | None |
| Typical B2B CPM | Bought on CPC, roughly $30-80 in competitive SaaS categories | ~$30-50 (LinkedIn ads cost) | ~$7-9 tech/SaaS (admanage.ai) |
| What it is uniquely good for | Catching a named buyer at the moment of search | Professional context and credibility | Cheap reach and frequency |
| Enriched match rate | 70-99% | 70-99% | 70-99% |
The reason to run the same list across all three is that they are not substitutes. Google catches demand you did not create. LinkedIn and Meta create it. This is the whole argument in contact-level advertising.
When does Google Customer Match not work for B2B?
Honest limits, because the play above is genuinely narrow.
→ You are under $50,000 lifetime Google spend. Targeting is unavailable. Exclusions and observation still work and are still worth doing, but the landmine play is not available to you today. Do not build a quarter's plan on it before checking your account's lifetime spend.
→ Nobody searches your category. Customer Match on Search only fires when someone searches. If your category has no search demand — a genuinely new product, or a term buyers do not know yet — you are waiting on a bell that never rings. Spend on the demand-creation channels instead.
→ Your CPCs are under $5. The whole economic argument is that filtering out non-buyers saves money proportional to CPC. At low CPC the enrichment and setup cost more than the waste you remove.
→ Your ICP is largely non-US. Person-level identity resolution is strongest in the US and degrades outside it. We publish what we have actually measured, including the bad numbers, on what we track today. Run a match test before committing budget.
→ You want to know which named person saw the ad. ContactLevel tracks ad clicks at the individual person level. It does not track ad impressions at the individual person level. If per-person impression counts are a hard requirement, that is a real gap and it is stated plainly on what we track today.
What does ContactLevel do here, and what does it not do?
Narrowly: it is the enrichment and sync layer between your CRM and the Customer Match list.
You pick or build an audience — from a CRM sync, a CSV, or contact search inside the platform. ContactLevel resolves each contact to the personal identifiers Google indexes, then keeps the Customer List in sync so it does not decay against the 540-day membership window. Match rates land at 70-99% rather than around 30% on a raw work-email export.
Pricing is published: $1,000/month for 10,000 net new contacts on Grow, $3,000/quarter for 50,000, $2,500/month for 30,000 on Grow, with a 14-day free trial and 10,000 free starter contacts (as of 2026-07-29, see pricing).
What it does not do: it does not get you enhanced access, it does not create search demand, and it does not report impressions per named person.
Go deeper.
The channel:
→ Google Ads for B2B SaaS — the four plays that fix a B2B Google account.
→ Google search inclusion — the step-by-step for the landmine play.
→ Google branded search exclusion — the play that works on basic access.
The mechanics:
→ Contact-level targeting — how identity enrichment works, platform by platform.
→ B2B match rates — what match rate actually measures and why the numbers differ per platform.
Compare:
→ B2B Facebook ads — the same identity problem on Meta, and the cheapest reach in B2B.
→ LinkedIn ads cost — what the professional-targeting premium actually costs.