Want ScalixAI to check whether your YouTube targeting finds buyers?
Key takeaways
YouTube targets behavior, not titles, so B2B precision comes from stacked proxy signals.
Demand Gen drops placement targeting; choose conversions or channel lists, not both.
Under $50K lifetime spend, Customer Match only excludes; it cannot target.
Your first five seconds filter buyers better than any audience setting can.
Judge targeting by work-email rate and SQL rate in your CRM.
YouTube ads targeting for B2B is the set of audience signals in Google Ads that puts your video in front of people who buy business software. YouTube has no job title, seniority, or company name targeting. So B2B targeting on YouTube relies on proxies: what buyers search for, which sites they visit, what your CRM already knows about them, and which people your creative filters in or out.
Most failing B2B YouTube accounts ScalixAI audits share one setup: broad "Business Professionals" affinity audiences, Optimized Targeting on, and a form fill as the goal. Views look cheap. Sales calls the leads students and freelancers.
Fyxer, an AI email assistant, started there. Its first two YouTube and Demand Gen tests spent about £23K for 7 conversions. After ScalixAI rebuilt the targeting, exclusions, and creative, the account hit £207 per conversion in January 2026, an 89% drop (Fyxer Demand Gen case study). This guide covers the targeting decisions behind it.
Can you target job titles or company size on YouTube?
No. YouTube does not let you target job titles, seniority, company names, or a list of target accounts. Google offers employment "detailed demographics" for company size and industry. Still, Google infers them from behavior; the bands are broad (one covers every company with 1 to 249 employees), and support varies by campaign type. ScalixAI does not build B2B targeting on them.
If your buyer is a VP of Security at a 200-person company, title precision lives on LinkedIn. Oneleet, a SOC 2 compliance platform, ran cold LinkedIn audiences split by role, seniority, title, and company size bands of 11 to 50, 51 to 200, and 201 to 1,000 employees. That level of control doesn't exist on YouTube (Oneleet LinkedIn Ads case study).
So YouTube has a different job in a B2B plan. LinkedIn decides who. YouTube reaches people showing buying behavior at lower CPMs, and Search captures them when they're ready. Our breakdown of Demand Gen vs Search covers how the two split the work.
[table]
What you want to target | Possible on YouTube? | The proxy that works
Job title or seniority | No | Search terms only that role types
Company size | Inferred and broad | Customer Match seeds from ICP-fit customers
Named target accounts (ABM) | No | Contact lists from those accounts via Customer Match
Competitor users | Indirectly | Custom segments from competitor URLs and brand terms
Category researchers | Yes | Custom segments from non-brand search terms
[/table]
Which YouTube targeting options actually work for B2B?
First-party data and custom segments work best for B2B. Broad affinity and generic in-market audiences work worst, because they reach people interested in business, not people buying software. Here is how ScalixAI ranks each signal for a B2B SaaS account.
[table]
Signal | What it catches | B2B reliability | Watch out for
Customer Match (CRM lists) | Known contacts, customers, lost deals | Highest | Work emails match poorly
Site and YouTube remarketing | People who visited or watched | High | Small pools early on
Custom segments: search terms | People who searched your terms on Google | High | Generic terms pull job seekers
Custom segments: URLs and apps | People who browse similar sites | Medium to high | A signal, not a visitor list
Lookalikes from customer seeds | People who resemble your buyers | Medium to high | Seed quality decides everything
In-market (e.g. business software) | Google's purchase-intent guess | Low to medium | Too broad for niche ICPs
Affinity (e.g. business professionals) | Long-term interests | Low | Mostly non-buyers
Detailed demographics: employment | Inferred company size and industry | Low | Broad, often wrong
[/table]
The closer a signal sits to your own data or a real search, the better it works.
Does your campaign type limit your targeting options?
Yes, and this catches many teams out. Demand Gen, the campaign type that replaced Video Action Campaigns for conversion goals, uses audiences only. It has no keyword, topic, or placement targeting, so you cannot tell it to run only on SaaS sales channels. According to Search Engine Land, you can see placement reports and add exclusions, but you don't pick channels to run on.
Video reach and video views campaigns still allow content targeting, including specific channels and videos. They optimize for impressions or views, though, not pipeline.
So founders face a choice. Want conversions? Run Demand Gen with audiences and exclusions. Want specific channels? Run a views campaign on hand-picked placements and use it to grow remarketing pools, not to book demos.
Demand Gen does let you choose surfaces. Google's channel controls are set at the ad group level, so you can limit serving to YouTube in-stream, in-feed, or Shorts and switch off Display or Gmail. For a B2B test aimed at YouTube, ScalixAI sets this before launch rather than letting "All Google channels" spend on mobile apps.
If you're new to how Demand Gen and YouTube fit together, start with our Demand Gen and YouTube ads playbook.
How do you build custom segments that find B2B buyers?
Build custom segments from the words and websites only your buyer uses at work. Use your converting non-brand search terms, competitor brand names, competitor pricing pages, and tools in your buyer's stack. Avoid generic terms like "sales software" or "AI tools", which also match students, job seekers, and creators.
Three rules ScalixAI applies:
- Start from terms that produced qualified leads, not clicks. Pull your Search terms report, join it to CRM outcomes, and use the terms that created opportunities. A term with 2,000 clicks and no pipeline will do the same thing on YouTube.
- Target what the buyer does at work, not what they sell. A tool for agency owners shouldn't target "marketing agency services", because that's what the agency's customers are searching. Target "agency client reporting tool" or "white label dashboard", which only the owner types.
- Add competitor URLs, then read them correctly. Adding a competitor's site tells Google to find people who browse sites like it. It is not a list of that competitor's visitors. Treat it as a signal and judge it on SQL rate.
Fyxer's prospecting segments used competitor intent, competitor landing pages, and non-brand search themes. US Prospecting delivered 48% of the flight's conversions at £320 CPA. GB Prospecting ran at £700, but it filled the pools that GB Lookalikes and Remarketing converted at £126 to £408. Our guide to Demand Gen audience structure covers how to split these into ad groups.
Why does first-party data matter more for B2B YouTube targeting?
Because your CRM knows who buys and Google's interest categories don't. Closed-won contacts, open opportunities, and churned customers tell Google more than any predefined audience. Check your access first, though.
Google's Customer Match policy gives every compliant account exclusions and observation. Full targeting needs at least 90 days of account history and more than USD 50,000 in lifetime spend. Many seed-stage AI companies can exclude customers on day one but can't target their own lists yet.
Two more things matter:
- Match rates. B2B lists of work emails often match poorly, because many people don't sign in to Google with their work email. Google recommends adding more match keys, such as phone numbers. Product signups that used personal emails usually match better.
- Seed by value, not volume. Fyxer's lookalikes were seeded from customer lists ranked by predicted lifetime value. GB Lookalikes became the best ad group in the account, with 524 conversions at £126 CPA. Seed with every free signup and Google finds more free signups.
Is Optimized Targeting bad for B2B?
Not by itself. Optimized Targeting lets Google look beyond your segments for whoever completes your goal most cheaply. If that goal is a form fill, it finds cheap form fillers: students, freelancers, and competitors.
The fix is the conversion signal, not the toggle. Send qualified events such as SQLs or activated workspaces back to Google before letting it expand. ScalixAI leaves Optimized Targeting off during the first learning month so it can see which segments drive results.
The goal also has to happen often enough to learn from. Fyxer's first test ran Maximize Conversions on a cold audience against a rare, deep goal, with a 30-day window shorter than its eight-week trial-to-paid cycle. It switched off after three weeks with 2 conversions. The rebuild kept a meaningful goal, moved the window to 60 days, and left bidding untouched for five months.
Is your creative doing the targeting for you?
Yes. On a platform with no job titles, the first five seconds of your ad decide who keeps watching. An ad that opens on a problem only your buyer has filters out everyone else before they cost you a click. An ad that entertains everyone gets cheap views from everyone.
Fyxer's data shows this clearly:
- Two GB image sets drew 178K clicks for 54 conversions. High CTR pulled in the wrong people.
- Brand stunts, party videos and polished parodies drew views but few signups.
- Problem-first workplace skits won. The best asset, "Wake Up Carl", ran 28 seconds at £118 CPA.
- 30-second cuts of winning ads were 82% cheaper per signup than the full-length versions.
For B2B, write a different hook for each buyer. "Your SDRs spend three hours a day on list building" speaks to a sales leader. "Your SOC 2 audit is in six weeks" speaks to a CTO.
Which exclusions should every B2B YouTube campaign have?
Exclusions do as much targeting work as inclusions. Without them, Smart Bidding finds the cheapest conversions: people who already pay you or never will. ScalixAI's pre-launch list:
- Existing customers and active trials, through Customer Match. This works on every account, even without full targeting access.
- Zero-value signups, such as personal-email signups that never activated.
- Channels you didn't mean to buy, using channel controls to switch off Display or Gmail when the test is about YouTube.
- Countries your sales team can't serve.
- Age bands your CRM shows never buy. Fyxer ran ages 25 to 44 only. Check your own CRM before copying that.
- Bad placements, reviewed weekly. Kids' content, gaming channels, and music videos show up in most B2B placement reports.
How do you know if your YouTube targeting is reaching B2B buyers?
Check the CRM, not the Google Ads dashboard. View rate and CPV say people watched, not that they could buy. Run these four checks on the last 30 days of YouTube-sourced leads:
- Work-email rate. What share used a company domain? A low share usually means consumer-heavy audiences.
- ICP fit. Enrich the leads and check company size and industry against your ICP.
- SQL rate by ad group. Compare each ad group with your Search campaigns. A big gap points to the audience or the offer.
- Top placements by spend. Open the placement report and read the top 50. You'll know quickly whether the money went to business content.
[table]
What you see | Likely cause | What to change
Cheap views, few site visits | Broad affinity or in-market audiences | Move budget to custom segments and first-party lists
Many signups, personal emails | Optimized targeting chasing a weak goal | Import qualified events, pause expansion
High CTR, low SQL rate | Creative appealing to everyone | Problem-first hooks that name the buyer's pain
Spend on apps and kids' content | Channels left on default | Set channel controls, add exclusions
Good leads, rising CPA | Pools saturating | Add new custom segments before raising budget
[/table]
Also watch branded search. Fyxer's monthly brand impressions rose from a 62K average to 169K in November 2025. After Demand Gen stopped, they fell to 51K by June 2026. Treat that as supporting evidence, not proof on its own.
Set your measurement window before judging targeting. Oneleet's LinkedIn dashboard reported a 0.36x return, while multi-touch attribution showed $1.63M in influenced closed-won revenue across 150 deals. Closes came months after first exposure: 1 in February, 36 in March, 47 in April. B2B targeting judged at day 14 almost always looks worse than it is.
Why should you change one targeting variable at a time?
Because otherwise you can't tell what worked. In spring 2026, a Fyxer relaunch changed ad groups, creative, and conversion goals at once while CPMs doubled. CPA rose to £1,776, and the cause couldn't be isolated. Change one audience, exclusion, or hook per test.
Want ScalixAI to check whether your YouTube targeting finds buyers?
ScalixAI will review your audience sources, Customer Match setup, channel controls, exclusions, conversion signal, and CRM lead quality, then show which segments produce pipeline and which produce cheap views. It's the same audit that came before the Fyxer rebuild.
Book a YouTube Targeting Audit
Still deciding? Read whether YouTube ads are worth it for B2B SaaS, compare B2B YouTube ads agencies, or see what clients say in our ScalixAI review.






