Most B2B accounts asking for value-based bidding do not have enough qualified conversions to run it, and the values they would feed it are guesses wearing a suit.
Value-based bidding is where B2B teams go when they get tired of counting leads. The logic is clean. Not every form fill is worth the same, so tell Google what each one is worth and let it chase the money instead of the volume. Then the account goes quiet, cost per qualified lead climbs, and nobody can explain why.
The reason is almost never the strategy itself. It is that the account was never eligible for it in the first place.
The short answer
If your Google Ads account produces roughly ten qualified conversions a month, do not run value-based bidding. Run a volume-based strategy pointed at the best qualified event your sales team actually maintains, and spend the next two quarters improving the quality and volume of that signal.
Google's own requirement for Target ROAS on Search and Shopping campaigns is at least 15 conversions in the past 30 days at the conversion tracking level. That is the eligibility floor for switching the strategy on. It is not the volume at which the strategy works well, and B2B teams read those two things as the same number constantly.
One naming change to expect if you go looking for the setting. Google started relabeling these strategies in June 2026, so "Maximize conversion value with a Target ROAS" now reads simply as "Target ROAS." The behavior did not change, only the name.
Ready for paid ads that pay off?
Book your free auditWhat value-based bidding is actually being asked to do
Volume-based bidding asks one question of every auction. How likely is this click to produce a conversion?
Value-based bidding asks a harder one. How likely is this click to produce a conversion, and how big will that conversion be? It is predicting a distribution, not a probability.
To predict a distribution, the model needs to have seen a distribution. It needs conversions that carry genuinely different values, enough of them to find the pattern, and it needs the pattern to be stable long enough to be worth learning.
Most mid-market B2B accounts fail all three tests at once. Ten to thirty qualified conversions a month, values assigned by a spreadsheet rather than observed from closed deals, and a sales motion that changes shape every time someone new joins the team.
Two or three static values is not value-based bidding
This is where it falls apart in practice. When a B2B team decides to "add values," what usually happens is a workshop where somebody assigns $500 to a contact form, $2,000 to a demo request, and $5,000 to a pricing-page inquiry.
Those are not values. Those are labels with dollar signs on them.
If a demo request is always worth exactly $2,000, then maximizing conversion value and maximizing conversions produce the same ranking of auctions, weighted slightly differently across three event types. You have added modeling complexity and a new way to be wrong, and you have bought nothing.
Google's own guidance on conversion values allows proxy values, and proxies are legitimate. But a proxy only earns its place when it varies with something real. A static tier does not.
→ Real variance looks like a lead value that moves with company size, deal stage, product line, or an actual closed-won amount imported from the CRM.
→ Fake variance looks like three fixed numbers chosen in a meeting, applied forever, and never checked against what those leads turned into.
The test is simple. Pull the last 90 days of qualified leads and look at the spread of the values you are sending. If the numbers cluster on three points, the bidding model is going to learn those three points and nothing else.
The sequencing that actually holds up
The order matters more than the strategy label, and most teams do it backwards. They pick a bid strategy first and then try to build a signal worthy of it.
1.) Move the qualified milestone onto the object your sales team maintains.
Not the object marketing wishes they maintained. If reps reliably update opportunity stage and nobody touches lead status, your qualified event is an opportunity, full stop. A signal that depends on a field nobody fills in is a signal that will quietly go dead. We cover the plumbing side of this in offline conversion imports for B2B.
2.) Set a threshold for what gets sent at all.
This is the single biggest improvement available to a low-volume account, and it costs nothing. Instead of assigning values to everything, decide which events are worth sending to Google and send only those. A lead-score cutoff, a minimum employee count, a business-email requirement. Binary gating on a clean definition beats a fake value model every time, because it sharpens the thing the algorithm is actually optimizing toward rather than adding noise on top of it.
3.) Bid on volume toward that gated event.
Maximize conversions, or Target CPA once you have stable data. The strategy is boring and that is the point. Our full argument on what to feed automated bidding lives in Smart Bidding for B2B.
4.) Consolidate so the signal lands in one place.
Ten qualified conversions a month split across six campaigns is one or two per campaign, which is not a learning signal, it is noise. Fewer campaigns and shared budgets is usually the precondition for any automated strategy working, as covered in B2B Google Ads campaign structure.
5.) Only then ask whether values would add anything.
By the time you have consistent qualified volume above the platform floor and real outcome data flowing back, the answer is often yes. It just was not yes in month two.
When value-based bidding is genuinely the right call in B2B
I am not against it. There are B2B accounts where it is clearly correct, and they share a shape.
→ Deal values genuinely vary by a factor of five or more, not by tiers but by actual outcome.
→ Closed-won or opportunity amounts flow back into the platform automatically from the CRM, not through a quarterly manual upload.
→ The account clears the platform floor comfortably, with 40 or more qualified conversions a month rather than 15.
→ Someone owns the value definition and reviews it against reality on a schedule, because a stale value model degrades silently.
→ The sales cycle is short enough that the values arriving today describe a market that still exists.
That last one does a lot of quiet damage in enterprise B2B. If your average cycle is nine months, the closed-won values feeding your bid model describe demand from three quarters ago. The model is optimizing for a market that has already moved.
How to switch without burning a quarter
If you do clear the bar, treat it as a change to the account's core economics rather than a setting change.
Run it as an experiment against the existing strategy rather than flipping the live campaign, and let it run long enough to produce a real verdict, which for most B2B accounts means at least six weeks rather than the two weeks people give it.
Expect volume to fall. That is the strategy working as designed, since it is buying fewer, more expensive clicks it believes are worth more. The question is not whether lead count dropped. It is whether qualified pipeline per dollar improved, which means you need pipeline reporting in place before you start, not after the first disappointing week.
And hold the target loosely at first. A Target ROAS set from a spreadsheet assumption rather than observed performance will either throttle the campaign to nothing or spend freely against a number it can hit easily. Start from what the account already produced, then move it.
What to do this week
Pull your last 90 days of conversions in Google Ads and count two things. How many qualified conversions the account produced per month, and how many distinct values are attached to them.
If the first number is under 15 or the second number is under five, you are not looking at a bidding problem. You are looking at a signal problem, and no bid strategy fixes a signal problem.
Fix the qualified event, gate what gets sent, consolidate the campaigns, and revisit values in two quarters when the account has earned them. If you want a second read on whether your account is actually at that threshold, that is the kind of thing our Google Ads team looks at first.
