ABM with Google Ads: when it actually makes sense

An industrial software client asked me to rebuild in Google Ads the list of 200 accounts his sales team was working in LinkedIn. Same logic, same ambition: pick the companies, show ads only to those companies, report by account. You can’t. Google Ads has no company targeting, no employer-industry targeting, no headcount targeting, and Google’s own catalogue of audience segment types makes that plain: affinity, custom segments, detailed demographics, life events, in-market, your data and similar segments. None of them is firmographic.

That doesn’t close the door on ABM. It moves the door. What Google contributes to a named-account programme isn’t the selection, which your CRM already did, but coverage of the demand those accounts create on their own when somebody inside starts searching. This post is about where that boundary sits, what you can build inside it, and what deal size makes the work worth doing. The full B2B strategy lives in my B2B lead generation guide; here I’m only talking targeting.

In 30 seconds:

• Google Ads doesn’t offer company targeting. None of its audience segment types is firmographic.
• Customer Match is the closest thing to ABM: it compares hashes against Google accounts, not against corporate domains.
• IP addresses can only be excluded, up to 500 per campaign. They’re useless for targeting.
• Google doesn’t restrict third-party trademarks as keywords, which makes competitor terms the most honest ABM lever on Search.
• To serve ads, a list needs at least 100 active users in the last 30 days.

Can you really do ABM in Google Ads?

Not in the strict sense a sales team means. ABM is closing a list of accounts and concentrating spend on them. In LinkedIn that’s a field in the campaign form. In Google Ads that field doesn’t exist, and I’ve spent years watching the expectation crash into the interface, usually in front of somebody who’d already promised it to the board.

The underlying difference is what each platform knows. LinkedIn knows where people work because people tell it and keep it current. Google knows what people search for. Two different signals, and neither converts into the other.

Accept the consequence early. In Google you don’t pick the account: it picks you when it searches. That’s a reactive position and it makes ex-outbound teams uncomfortable, and it’s exactly why the channel performs when it performs, because whoever types “predictive maintenance software pricing” has a problem today rather than two quarters from now. So the useful question isn’t whether Google Ads does ABM. It’s which layers of your programme it can cover. If you’re starting from zero on the channel, the base approach is in Google Ads for B2B.

What can you target and what can’t you?

Four real levers. And a long list of things people assume without checking. I’ll start with the whole map, because most arguments about ABM in Google Ads get settled by looking at it for a minute.

What doesn’t exist deserves saying out loud. There’s no email-domain targeting, no job title, no company size. No IP targeting either: Google lets you exclude up to 500 IP addresses per campaign, and that feature pulls ads away from specific places, it never points them at those places.

The lever that pays best is usually the least ABM of the lot: competitor terms. Google’s trademark policy is explicit and puts the use of trademarks as keywords among what Google Ads won’t restrict. You can bid on the name of the vendor your target accounts use today, on the single condition that you don’t write it into the ad copy. Nearly everyone misses that distinction the first time round.

One warning is left. If you apply your list in targeting mode to a Search campaign that already performs, traffic collapses; the mechanics are in audience layering in Google Ads.

Why does Customer Match underperform in B2B?

Because the match isn’t against the company. It’s against the person, and specifically against their Google account. When you upload a file, Google turns the data into codes using the SHA256 algorithm and compares those codes with the ones for Google accounts, which are hashed too. The same page says it plainly: customer information often doesn’t match.

In B2C that works reasonably well, because the email your customer gave you is often the one behind their Gmail. In B2B the email in your CRM is a work address. It’ll match if that domain runs on Google Workspace, or if the person registered that address with a Google account at some point. If the company runs Microsoft 365 and the contact has never touched a Google product with that address, there’s nothing to pair. It isn’t a fault in your file. It’s the mechanism, and in my experience it’s where most B2B ABM plans quietly come apart.

Hence what I recommend before building anything: upload the list and read the match rate, which Google defines as the percentage of your uploaded customer list that can be connected to Google users. If it matches poorly, you’ve saved yourself a month of expectations.

There are also requirements that don’t depend on you. Google asks for a good policy compliance history and a good payment history, and reserves full targeting with bid adjustments for accounts with more than 90 days of history and more than $50,000 in total spend. If you’ve just opened the account, this lever isn’t on the table.

Membership expires, too. Google caps list membership at 540 days and requires that at least 100 members have been added or updated in the last 540 days for the list to stay eligible. A CRM export uploaded once and forgotten switches itself off. Wire the upload into the same pipeline you use to send offline conversions from your CRM.

What account size justifies building it?

There’s a technical floor and an economic one. The technical one kills more projects. To serve ads with a data segment, the list must have a minimum of 100 active visitors or users within the last 30 days on Search, Display and YouTube. Look at what that means with a 50-account target list: you’d need two active contacts per account, all with matching emails, all active in the same month. It rarely happens.

The economic floor is softer but just as real. Maintaining this layer costs hours: cleaning the CRM, automating the upload, building landing pages, checking match rates, reading results on tiny volumes. Those hours pay for themselves when a contract is worth tens of thousands and the cycle runs for months. On a three-figure deal, they don’t. Is it worth an afternoon a month for an audience of 140 people? It depends entirely on what closing one of them is worth.

My working rule allows exceptions: under roughly 300 identified contacts in the CRM I’d rather put that effort into competitor keywords and the landing page. Above that it starts to earn its place. Above several thousand it isn’t ABM any more, it’s ordinary remarketing. To sanity-check the numbers, the B2B cost per lead framework beats any industry benchmark.

How do I build the ABM layer when it pays?

In this order, and the order matters because each step depends on the previous one working:

1. Export and clean the CRM list. Email and phone, not email alone: Google recommends adding as many match keys as possible, and in B2B the mobile number rescues contacts the work address loses.
2. Upload it and read the match rate. That’s the real decision point. If it matches poorly, stop here.
3. Apply it in observation, never in targeting, on the Search campaigns that already work.
4. Build competitor campaigns separately, with their own budget and ads that never name the rival brand.
5. Create landing pages for the target accounts and don’t link them from the menu. That way remarketing from that page only picks up people your outbound brought in. How to structure them is in B2B landing pages.
6. Close the loop with offline conversions. Without that you’re optimising for form fills, not opportunities.
7. Review at 90 days, not 15. On small volumes, a weekly read is noise.

Step five has produced the most for me and it’s the one almost nobody does. A URL that only sales hands out by email turns that visitor list into the cleanest approximation of a named-account audience you’ll get. You don’t know which company is behind it. You know how they arrived. The mechanics are in my Google Ads remarketing guide.

Which mistakes do I keep seeing?

Three, always the same ones.

The first, and I’ve seen it in nearly every audit, is treating the customer list as if it were the company list. It isn’t. It’s a list of people who may or may not show up, and any plan assuming full coverage of the 200 accounts is built on sand. The second is applying strict targeting to Search campaigns that were already delivering: volume drops, the algorithm loses signal, and two weeks later somebody asks about the leads.

The third is measuring this the way you’d measure an ecommerce account. With audiences of a few hundred people and long cycles, weekly CPA says nothing. What does say something is how many target accounts entered the pipeline during the quarter, and you only know that if your lead grading is connected to the ad that brought them. The model I use is in B2B lead scoring.

Frequently asked questions

Can you target specific companies in Google Ads?

No. Google Ads offers no firmographic audience segment: no company, no employer industry, no headcount. The closest thing is Customer Match, which reaches specific people from your CRM when their email or phone matches a Google account, plus remarketing on landing pages only your sales team distributes.

Can you target an IP address to reach an office?

No. Google Ads only lets you exclude IP addresses, up to 500 per campaign, and the feature exists to pull ads away from specific locations. The inverse operation doesn’t exist. I use it to take the client’s own office out of the impression count, which is account hygiene rather than an ABM tactic.

How many contacts do I need for the list to work?

To serve ads, a data segment needs at least 100 active visitors or users within the last 30 days on Search, Display and YouTube. In practice you need considerably more, because only part of your work email addresses will match Google accounts. Below a few hundred CRM contacts, in most cases it doesn’t hold up.

Can I bid on my competitors’ names?

Yes. Google’s trademark policy places the use of trademarks as keywords among what Google Ads won’t restrict. What it does restrict is using a rival’s trademark in the ad text. It’s one of the most useful levers in B2B, though clicks tend to be expensive and you should measure them against opportunities, not form fills.

Is ABM in Google Ads worth it if I already spend on LinkedIn?

It depends on deal size and on how much search volume your category has. LinkedIn covers account selection, which Google can’t do. Google covers the moment somebody inside those accounts looks for a solution, which LinkedIn never sees. They sit badly together if you expect the same thing from both, and well if you give each one its stage.

ABM in Google Ads is a layer, not a channel

Building this a few times taught me that the value sits in accepting the limit early. Google isn’t going to hand you the company list. It hands you the search, a later and hotter signal, and lets you reinforce it with what you already know about your customers. Anyone who walks in expecting LinkedIn with a different logo ends up blaming the channel, and I’ve had to talk more than one team down from that.

With a high deal value and a wide contact base, the layer holds up: you catch the accounts that activate on their own, you reinforce on people who already know you, and you don’t leave your competitors’ terms to your competitors. With a low deal value or a short list, the same build costs more hours than it returns. Saying so is more honest than invoicing it.

If you’d like me to look at your case before you sink time into this, tell me how many target accounts you have and what closing one is worth, book 30 minutes of consulting and I’ll tell you whether the layer holds up or whether your money works harder somewhere else.