The average B2B Google Ads account wastes between 40% and 60% of its budget on clicks that were never going to convert. Not because the keywords were wrong. Because there was no audience layer telling Google who you actually want clicking those ads.
Keywords tell Google what someone is searching. Audiences tell Google who’s doing the searching. Run B2B campaigns without both, and you’re handing your budget to interns, students, and competitors doing research — not the VP of Operations who needs your software by Q3.
- Keyword targeting alone is not enough in B2B — you must layer audience signals to qualify who sees your ads, not just what triggers them.
- Customer Match is your highest-intent B2B audience and most accounts use it wrong or not at all.
- In-market audiences work in B2B, but only the right segments — and you have to test aggressively to find them.
- Audience layering (stacking multiple signals on one campaign) is how you move CPL down without cutting reach.
- Observation mode is your friend before you bid adjust — rushing to target mode kills scale before you have the data to justify it.
Why Most B2B Google Ads Accounts Are Structurally Broken From Day One
Here’s what a typical B2B account setup looks like: someone picks 50–200 keywords, writes 3 RSAs per ad group, sets a target CPA, and calls it a day. Maybe they add remarketing. Maybe.
The problem isn’t the keywords. It’s that those keywords get served to anyone who types the right phrase. “Enterprise project management software” gets clicked by a Fortune 500 IT director — and by a university student writing a paper on SaaS business models. Same keyword. Wildly different commercial value.
Google’s algorithm, left to its own devices, will chase volume. It optimizes toward the conversion signal you give it. If your lead form is the signal, it will find leads — including garbage ones. Audience targeting is how you teach Google to chase the right volume.
This isn’t a nice-to-have for B2B. It’s load-bearing infrastructure.
Start With What You Already Have: Customer Match Is Your Secret Weapon
Customer Match lets you upload a list of email addresses — existing customers, churned accounts, target prospect lists, conference attendees, CRM contacts — and Google matches them to signed-in Google users. Then you can either target them directly or use them as a seed for similar audiences.
Most B2B teams either don’t use it at all, or they upload one list once and forget about it. That’s leaving your most powerful signal completely idle.
How to Actually Use Customer Match in B2B
Suppress your existing customers. The fastest way to lower your CPL is to stop paying for clicks from people who already pay you. Upload your full customer list as a negative audience on acquisition campaigns. This alone will move your numbers.
Create a “Dream 100” list. Pull your top 100–200 target accounts from your CRM or from a tool like ZoomInfo. Upload the known business email addresses for decision-makers at those companies. Run a dedicated campaign or ad group with messaging built for cold-but-aware prospects. This is account-based marketing running inside Google Ads.
Build a churned customer re-engagement segment. These are people who knew your product well enough to pay for it. They’re far easier to win back than cold prospects. Separate list, separate messaging, separate bid strategy.
One caveat: Customer Match requires a minimum list size (typically 1,000 matched users) before it activates, and match rates for B2B are usually 30–50% because personal Gmail addresses match better than corporate ones. Upload more contacts than you think you need. And refresh your lists every 30–60 days — stale data degrades performance quietly.
In-Market Audiences for B2B: They Work, But Not the Way You Think
The conventional wisdom is that in-market audiences are mostly a B2C play — people shopping for cars, planning vacations, comparing insurance. That’s wrong, or at least incomplete.
Google has dozens of B2B-relevant in-market segments. Business Services, Enterprise Software, Shipping & Logistics, Financial Services & Insurance, HR & Staffing — the list is longer than most people realize. These segments identify users whose recent search behavior and content consumption signals active purchase research.
The honest caveat: in-market audiences in B2B are noisier than in B2C. The signal is real but impure. Someone flagged as in-market for “CRM software” might be a genuine VP of Sales evaluating vendors — or a marketer writing a comparison article. You don’t know yet.
The Right Way to Test In-Market Audiences in B2B
Start in observation mode, not target mode. Add every plausible in-market segment to your existing campaigns as observation audiences. Run it for 30–45 days. Then pull the audience performance report and look at two things: conversion rate by audience segment, and CPA by audience segment.
You’ll almost always find that 2–3 segments dramatically outperform your campaign average. Those are the segments you apply positive bid adjustments to — typically +15% to +40% depending on how strong the signal is. The segments that underperform, you either ignore or apply a negative adjustment.
Only after you’ve validated performance do you consider moving to target mode. Target mode restricts your reach to only the selected audiences. In B2B, where total audience pools are already small, that can hurt scale badly if you do it too early.
Audience Layering: Where the Real Efficiency Gains Live
Audience layering is the practice of stacking multiple audience signals on a single campaign or ad group so that your ads show (or bid higher) only when someone meets more than one criteria simultaneously.
The logic is simple: someone searching “cloud security software” is a decent lead. Someone searching “cloud security software” who is also in-market for enterprise IT solutions AND has visited your pricing page is a completely different conversation. You should bid very differently on those two people.
A Layering Framework That Actually Works for B2B
Think of your audience signals in three tiers:
Tier 1 — Intent signals: What they’re actively searching (keywords). This is your base layer. It’s table stakes.
Tier 2 — Context signals: In-market segments, custom intent audiences (users who’ve searched specific competitor terms or industry phrases), and similar audiences built from your Customer Match lists. These tell you something about their professional context.
Tier 3 — Relationship signals: People who’ve visited your site, engaged with your YouTube content, been on your customer list, or been uploaded via Customer Match. These are your warmest signals.
The layering move is to bid differently based on how many tiers a user hits at once. Someone hitting Tier 1 only? Bid normally. Tier 1 + Tier 2? Bid up 20%. Tier 1 + Tier 2 + Tier 3? Bid up 40–60% — that’s your highest-intent prospect and you can afford to pay significantly more to win that impression.
You implement this through bid adjustments in manual or enhanced CPC campaigns, or by building separate ad groups or campaigns for each tier with different target CPA/ROAS goals in Smart Bidding campaigns.
Custom Segments: Build Your Own In-Market Audiences When Google’s Don’t Fit
Google’s predefined in-market segments were built for scale across millions of advertisers. They’re broad by design. For B2B advertisers with a specific ICP — say, CFOs at mid-market manufacturing companies — those segments will rarely carve out your exact buyer.
Custom segments (formerly called custom intent or custom affinity) let you build your own audience definition based on:
- Search terms people have recently Googled (you define the keywords)
- URLs people have recently visited (competitor sites, industry publications, review sites like G2 or Capterra)
- Apps people use
For B2B Google Ads, the URL-based approach is powerful and underused. Think about who your buyer is and where they live online. If you sell ERP software to manufacturers, your ICP probably reads Industry Week, visits SAP.com and Oracle.com to compare options, and uses G2 to research alternatives. Build a custom segment targeting users who’ve recently visited those URLs.
You’re essentially telling Google: “Find me people who have been hanging out in spaces where my buyer hangs out.” It’s not perfect, but it’s a dramatically sharper audience than “anyone who searched ‘ERP software’ in the last 30 days.”
Layer this on top of your keyword campaigns in observation mode first. If it performs, bid it up. If it doesn’t, scratch it and try different seed URLs.
Remarketing in B2B: Stop Treating All Site Visitors the Same
Your remarketing list is not one audience. Treating it as one is one of the most expensive mistakes in B2B paid search.
Someone who bounced off your homepage after 8 seconds is not the same as someone who spent 4 minutes on your pricing page, visited two case studies, and started a demo request form before abandoning. Those are two completely different intent signals, and they deserve completely different ads, bids, and messaging.
The Remarketing Segments Every B2B Account Should Have
Homepage bouncers (0–30 seconds): Low intent. Don’t waste high bids here. Either exclude them or run very conservative bids focused on brand-level keywords only.
Content/blog visitors: Early-stage awareness. Good candidates for thought leadership ads, guide downloads, and webinar invitations. Not yet ready for “Book a Demo” CTAs.
Product/solution page visitors: Mid-funnel. These people are evaluating options. Hit them with comparison content, social proof (customer logos, case study snippets), and differentiation messaging.
Pricing page visitors: High intent. They looked at the number. They’re comparing. Bid aggressively here — these are your best non-Customer-Match leads. Test urgency-based offers: limited-time demos, ROI calculators, direct sales outreach offers.
Abandoned form starts: The highest-intent segment you have outside of Customer Match. Someone filled in their name and email and then stopped. A specific ad sequence for this audience — maybe with a softer ask like a content download — can recover a meaningful percentage of these.
Build these as separate audiences in Google Ads (or import from GA4 if you have it connected), then layer them with time windows. Someone who visited your pricing page yesterday is hotter than someone who visited six months ago. Decay your bids accordingly — aggressively high in days 1–7, tapering through day 30, pulling back sharply after day 60.
Frequently Asked Questions
Does audience targeting work differently on the Google Search Network vs. Display Network for B2B?
Yes, significantly. On Search, audiences modify who you bid on within a keyword-triggered auction — your ad still only shows when someone searches your keyword, but you can bid more for users who also match your audience. On Display, audiences replace keyword targeting — you’re finding people across the web based purely on who they are. For B2B, Search with audience layering is almost always more efficient than Display targeting, which tends to generate a lot of low-quality impressions and clicks. Start with Search, prove your audiences work, then consider Display for remarketing and awareness only.
What’s a realistic match rate for B2B Customer Match lists?
Expect 20–50% match rates for B2B lists. Corporate email addresses (john.smith@companyname.com) match at lower rates than consumer Gmail or Yahoo addresses because users are less likely to sign into Google with their work email. To improve match rates, include phone numbers and physical addresses alongside email in your upload. And upload significantly more records than you think you need — a list of 500 contacts might only activate 150–200 matched users, which may fall below minimum thresholds.
How long should I run audiences in observation mode before acting on the data?
Minimum 30 days, ideally 45–60 days if your conversion volume is low (under 50 conversions per month). You need statistically meaningful data before you apply bid adjustments. Pulling the data after two weeks and seeing one audience segment with a 0% conversion rate tells you almost nothing — it might have had 12 impressions. Wait for volume, then act.
Can I use LinkedIn audience data inside Google Ads for B2B?
Not directly — LinkedIn and Google don’t have a native data-sharing integration for audience targeting. However, you can export contact lists from LinkedIn Campaign Manager (from Lead Gen Forms or matched audiences) and upload them to Google Ads as Customer Match lists. It’s a manual bridge, but it works. Some B2B teams run LinkedIn ads specifically to build warm contact lists they then retarget more cheaply on Google.
Should I use Similar Audiences (Similar Segments) for B2B prospecting?
With caution. Google’s Similar Segments can expand your reach by finding users who look like your existing Customer Match or remarketing lists. For B2B, the challenge is that “similar to your website visitors” can be a very noisy signal — your site attracts competitors, job seekers, and students alongside actual buyers. If you use Similar Segments, do it in observation mode first, and watch your lead quality carefully. A cheaper CPL means nothing if the leads don’t qualify downstream.
What’s the minimum budget to make B2B audience targeting worthwhile?
There’s no hard floor, but realistically you need enough volume to generate data. If your campaigns are getting fewer than 500 impressions a week per ad group, audience signals won’t have enough to work with. The more meaningful threshold is conversion volume — you need at least 30–50 conversions in a period before segmenting by audience performance becomes statistically reliable. For most B2B advertisers, that means a minimum of $3,000–$5,000/month in spend before audience layering starts showing clear ROI improvements.
Here’s an honest checklist. If your current Google Ads setup doesn’t have at least these four things in place, you’re leaving serious money on the table:
- Customer Match lists uploaded, refreshed, and actively used for suppression and targeting
- At least 5–7 distinct remarketing segments based on page type and intent depth — not one giant “all visitors” list
- In-market and custom segments running in observation mode with documented performance benchmarks
- Bid adjustments (or separate campaigns) that reflect actual audience intent tiers — not a flat bid across all users
If your agency’s monthly report doesn’t mention any of these by name, ask them directly what your audience strategy is. A vague answer is your answer. If you want a second set of eyes on your account structure — no pitch, just a real audit — reach out here and we’ll tell you exactly what we see.