In October 2026, Google added qualifying responses to lead forms — you can now mark specific answers to your custom questions as "qualifying," and a submission that contains one is counted under an auto-created conversion action named "Lead form - Response qualified (Google-hosted)." That gives Smart Bidding a same-session quality signal it never had before. The catch: it measures what a prospect says about themselves, not whether sales will touch the lead — so it is faster than CRM offline conversions, and shallower.
This is for revenue-accountable marketing leaders at Series A to Series C B2B SaaS companies who already run lead forms and are deciding what signal to let the algorithm chase. The short answer: use qualifying responses to shorten the learning period, but anchor pipeline-grade bidding on the sales-qualified lead (SQL) and closed-won signals from your CRM. Below is how the two signals differ, when each is the right primary, and how to wire them together without teaching Smart Bidding to buy junk.
What actually changed in October 2026
Lead forms have let you add custom questions for years. What is new, per Search Engine Roundtable's 2026-10-16 report, is that each answer to a custom question can now be flagged as qualifying. When you build the form under Assets → Lead form → Custom questions, a question can carry up to 12 answers, and you check which ones count as qualifying. A submission containing a qualifying answer is tagged and, per Google's help documentation, counted under the auto-created "Lead form - Response qualified (Google-hosted)" conversion action. The feature is documented as Search-only at launch.
The important part for a B2B SaaS advertiser is not the UI — it is that this creates a standard conversion action you can set as primary and bid toward. In other words, Google has shipped a native, in-form pre-qualification signal that sits structurally between a raw form fill and a CRM-validated SQL. The question is whether it deserves to be your bidding target.
The two signals, and why they are not interchangeable
Every lead-gen Smart Bidding strategy optimizes toward a conversion action, and the depth of that action decides what the algorithm learns to buy. There are now three depths available to a B2B SaaS account, and they differ on two axes that matter: latency (how fast the signal is available to bid on) and truth (how well it predicts an actual sales-qualified lead).
| Bidding signal | Latency | Truth (predicts SQL?) | Gameable by prospect? |
|---|---|---|---|
| Raw form fill / lead | Instant | Weak | Yes — zero friction |
| Qualifying response (new) | Instant (same session) | Moderate — only as good as the question | Partly — self-reported |
| CRM offline conversion (SQL / closed-won) | Days to weeks | Strong — sales-validated | No — human-reviewed |
A qualifying response is a real improvement over bidding to raw form fills, because a self-selected "500+ employees" or "evaluating now" answer correlates with ICP fit better than a bare submit does. But it is still self-reported and still fires in-session, so it carries none of the downstream truth that a sales-qualified lead (SQL) — a lead sales has agreed to work — or a closed-won deal carries. If your qualifying question is soft, you have simply moved the junk problem one step downstream.
When each signal should be your primary
The decision is mostly about data volume and sales-cycle length. If your account generates fewer than the roughly 30–50 conversions per campaign per month that Smart Bidding needs to learn, CRM offline conversions alone starve the algorithm — SQLs are too sparse and too delayed to train on. That is exactly where qualifying responses earn their place: a denser, faster proxy that keeps bidding out of the learning period while CRM data accumulates. We cover the volume threshold in the Smart Bidding learning-period post.
- Use qualifying responses as primary when lead volume is low, the sales cycle is long, and SQL data is too sparse to bid on directly — and when you have a genuinely discriminating qualifying question tied to ICP firmographics.
- Use CRM offline conversions as primary once you have enough SQL/closed-won volume to train bidding, or whenever the qualifying answer has proven a poor proxy for real sales acceptance.
- Run both in nearly every mature account: qualifying responses to shorten learning and stabilize bidding, CRM offline conversions (via enhanced conversions for leads) as the signal you steer toward over time. See the offline conversion stack for the plumbing.
Picking a qualifying question that actually predicts pipeline
A qualifying response is only as good as the answer you flag, so the whole value of the feature rides on question design. The goal is a question where the qualifying answer is honestly selected only by people who resemble your closed-won customers — a firmographic or intent filter, not a vanity checkbox. "Are you interested in a demo?" is worthless because everyone says yes. "How many employees?" with only the in-ICP bands marked qualifying is useful because it splits the in-market from the tire-kickers.
Before you promote the qualifying-response conversion to primary, validate it against reality: pull the submissions that contained the qualifying answer and check their actual SQL rate in HubSpot or Salesforce against submissions that did not. If the qualifying cohort does not convert to SQL at a meaningfully higher rate, the question is not discriminating and you are about to teach Smart Bidding to buy a slightly fancier form fill. This is the same lead-quality discipline described in conversion value adjustments for lead quality and optimizing for SQLs, not leads.
Wiring both signals together
The durable setup is a two-layer conversion structure. Layer one is the qualifying response, counted as a primary conversion while volume is thin, so Smart Bidding has a dense signal to learn on. Layer two is the CRM offline conversion — SQL and closed-won pushed back through enhanced conversions for leads — which becomes the primary signal once enough pipeline data exists to train on. You transition the primary from layer one to layer two as data matures; you do not run both as primary forever, because competing primaries confuse the bidding objective.
One caution specific to this rollout: qualifying responses are Search-only today. If a meaningful share of your demand runs through Demand Gen or Performance Max, you cannot lean on this signal there, and CRM offline conversions remain your only quality feed for those campaigns. That also means a qualifying-response conversion and a CRM SQL conversion can double-count the same lead if you are not careful about which is marked primary per campaign — keep the counting clean so your reported cost per SQL stays honest. For the related "non-biddable goals" mechanic, see journey-aware bidding.
The Two Spouts view
Qualifying responses are a genuinely useful addition, and we will turn them on for most funded B2B SaaS accounts — but as a learning-period crutch, not as the destination. The failure mode we expect to see over the next two quarters is teams flipping the "Lead form - Response qualified" action to primary, watching cost per "qualified lead" look great, and quietly recreating the exact junk-lead problem the feature was supposed to solve — because the qualifying question was soft and nobody checked it against SQL rate. A faster signal that you have not validated is just a faster way to buy the wrong leads.
The pipeline-grade answer has not changed: Smart Bidding should steer toward what sales accepts and what closes, which only your CRM knows. Use qualifying responses to get there faster, not to avoid going there at all. If you are not sure whether your account is bidding to a real quality signal or a cosmetic one, a Google Ads audit will trace exactly what your campaigns optimize toward — start with the free 10-point audit, or see how we run paid search for funded SaaS teams.