Most B2B SaaS accounts hand Google Ads a single, flat conversion: a form fill, a trial signup, a demo request. Every one counts as exactly one, so Smart Bidding optimizes to produce as many of them as cheaply as possible. The predictable result is a campaign that gets very good at generating the lowest-value version of a conversion — gated PDF downloads, tire-kicker trials, leads sales never calls back — because that is literally what you asked it to maximize.
A conversion value ladder fixes the instruction. Instead of telling the algorithm "get me conversions," you tell it "a closed opportunity is worth 200 times a content download, and here are the values in between." Fed into value-based bidding, that turns Google Ads from a lead-volume machine into a pipeline-value machine. This is a vendor-neutral, evergreen playbook for building that ladder and wiring it in without starving the bidder of data.
Why flat conversions quietly fail
When every conversion carries the same implicit value, Smart Bidding has no way to tell a future customer from a future unsubscribe. It bids the same for both, and because low-intent conversions are cheaper and more plentiful, the algorithm drifts toward them over time. You see a falling cost per lead and a rising lead count, and it looks like progress — right up until sales reports that the pipeline did not move. The metric improved because the definition of success was wrong.
This failure mode is sharpest in channels built for volume. Performance Max, in particular, "tends to over-index on low-quality conversions (content downloads, repeat visitors) because those are easier to generate at volume," as one 2026 SaaS playbook puts it (groas.ai). The value ladder is the antidote: when a content download is worth $5 and a demo is worth $500, the bidder stops chasing the cheap thing on its own. For the broader question of which events to even count, start with conversion tracking for SaaS.
Step 1: map your funnel stages
Before assigning numbers, list the stages a prospect actually passes through, from first touch to revenue. A typical B2B SaaS ladder has four to five rungs: a top-of-funnel micro-conversion (content download, newsletter), a product or sales entry point (free trial or demo request), a sales-qualified lead (SQL), an opportunity, and finally closed-won. Not every account needs all five — a pure product-led company may collapse demo and SQL — but the ladder should mirror how your revenue is genuinely created, not an idealized diagram.
Be honest about which of these are leading indicators and which are vanity. A "request more info" click that never correlates with revenue does not belong on the ladder at any value, because giving it even a small weight teaches the bidder to manufacture it. The discipline of this step is subtraction as much as ordering: keep only the stages that predict money, and arrange them so each rung is meaningfully closer to cash than the one below it. This pairs naturally with structuring the account by funnel and intent tiers so each campaign feeds clean signal.
Step 2: assign revenue-weighted values
Now put dollar figures on each rung, derived from your own funnel math rather than guesses. The method: take average contract value, multiply by the historical close rate from each stage, and use the result as that stage's value. If demos close at 20% against a $10,000 ACV, a demo is worth roughly $2,000 as a signal; if trials close at 5%, a trial is worth about $500. A widely-cited 2026 model lands in the same neighborhood — "trial = $50, demo = $500, SQL = $2,000, enterprise opportunity = $10,000+" — the exact figures will differ for your business, but the order of magnitude between rungs is the part that drives bidding behavior.
Two cautions. First, the ratios matter more than the absolute numbers: if a demo is ten times a trial in reality, your values must encode that 10x or the bidder will misprice the funnel. Second, refresh these values as cohorts mature. Close rates and contract values drift, and a ladder built on last year's economics slowly lies to the algorithm. Treat the values as a living input reviewed quarterly, not a one-time setup. If you sell across very different segments, consider tiering values by ACV band so an enterprise demo and an SMB demo are not flattened into one number.
Step 3: close the loop with offline data
The top of the ladder lives on your website and can carry values immediately — a trial signup can fire with a $500 value the moment it happens. But the rungs that matter most for accuracy, SQL and opportunity and closed-won, live in your CRM and only become known days or weeks later. To bid on them, you have to send them back to Google Ads with their values through offline conversion import, which in 2026 means the Data Manager API. Without that feedback loop, value-based bidding optimizes to the cheapest on-site proxy and never learns what actually closed.
This is the step most accounts skip, and it is why so many value-based bidding rollouts underperform. Sources are blunt about the cost: value signals only work "when paired with offline conversion tracking, ICP-quality signal feedback, and value-based bidding. Without those, PMax burns 40–60% of budget" (groas.ai). The plumbing is non-trivial but well-documented — see our walkthrough on migrating SaaS conversion uploads to the Data Manager API — and it is the difference between a ladder that informs bidding and one that just decorates a slide.
Step 4: feed it into value-based bidding
With values assigned and offline data flowing, switch the relevant campaigns from a conversion-count strategy to a value strategy — Maximize Conversion Value, optionally with a target ROAS once you have enough data to set one credibly. The bidder will now weight its bids toward the rungs that carry the most value, spending more to win a click likely to produce a demo than one likely to produce a download. The choice between an uncapped value strategy and a target-ROAS one depends on volume and stability; our guide to Maximize Conversion Value vs tCPA breaks down when each fits.
Give the switch room to work. Value-based bidding needs conversion volume and a few weeks to recalibrate, and reported numbers often look worse before they look better as the algorithm stops chasing cheap conversions and starts paying up for valuable ones. Judge it on downstream pipeline and CAC, not on cost per lead — the whole point of the ladder is that cost per lead is no longer the goal. Watch attribution carefully during the transition, because the model you use changes which touchpoints get credit for those higher-value conversions; our attribution models guide covers the trade-offs.
Common mistakes that break the ladder
The most frequent error is building the ladder without the offline loop, so the deep rungs are theoretical and the bidder still only sees on-site events. The second is making the rungs too close together — a ladder where a download is worth $40 and a demo is worth $50 barely changes behavior, because the algorithm sees almost the same number for both. Spread the values to reflect real economics; timid ratios produce timid results. The third is forgetting to maintain the values as your funnel economics shift, leaving the bidder optimizing to a stale picture of your business.
A subtler trap is over-counting early-stage volume. If your top-of-funnel event fires far more often than your valuable events, even a small per-event value can sum to a large share of total tracked value, quietly pulling spend upward in the funnel. Keep micro-conversion values genuinely small, or exclude them from the bidding signal entirely and use them only for diagnostics. The ladder works because the gaps between rungs are honest — protect those gaps, keep the offline feedback flowing, and value-based bidding will spend your budget where the revenue is.