If forms are filling up but sales are not, the problem is usually not your sales team. It is the signal you send back to the ad platform. Lead quality is the probability that the people your ads bring in actually become customers, and form count does not measure it. Feed Google Ads or Meta nothing but 'form submitted' and the algorithm becomes excellent at finding form fillers, not buyers. The fix is to define the qualified stage as its own conversion action and upload that outcome back to the platform.
What is lead quality?
Lead quality is the rate at which records from an ad channel move through your sales process and turn into revenue. The unit of measurement is not how many forms arrive, but how many of them reach the stage sales calls qualified. Industry benchmarks put the share of marketing-qualified leads that sales also considers qualified at roughly 13 percent. In other words, the volume report and the revenue report rarely tell the same story.
- Raw record: everyone who submitted the form. This is the number your ad dashboard calls a conversion.
- Qualified lead: someone whose budget, authority and need match what you sell. Only your sales team or CRM knows this.
- Won deal: the moment the contract is signed and money moves. In long sales cycles this happens weeks or months after the form.
What quality signal are you feeding the platform?
Smart Bidding systems only learn from the event you report back. If your conversion action is 'form submission', the system hunts for people likely to submit forms. That group includes content downloaders, people typing fake numbers and buyers with no purchasing authority. Upload the qualified stage and its value instead, and the same algorithm starts looking for audiences that resemble your actual customers. Without feedback, optimisation locks onto volume.
- No signal: only form submissions count. The platform maximises volume and quality is left to chance.
- Stage signal: 'qualified' and 'won' are defined as separate conversion actions and uploaded back.
- Value signal: each stage gets a monetary value or score, so bidding optimises for value rather than count.
- Identity match: the CRM record is matched to the ad click through a click identifier or hashed email. A low match rate means a weak signal.
How do you define a qualified lead as a conversion action?
Short answer: agree on a one-sentence definition of 'qualified' with sales, bind it to a single CRM field, then upload that field to the platform as its own conversion action. A fuzzy definition produces fuzzy data. In practice the best balance is to make the qualified lead your primary conversion goal and keep won deals as a secondary observation goal, because closed-won data is far too sparse in most businesses for bidding to learn from.
- Sit down with sales and write the qualified definition: which budget, which region, which need. Leave nothing open to interpretation.
- Bind that definition to a single status field in the CRM and store the click identifier (GCLID, wbraid or equivalent) on every record.
- Create separate conversion actions in the ad account for 'qualified lead' and 'won deal'.
- Upload records marked qualified on a regular schedule, ideally daily. The longer the delay, the slower the learning.
- Move the primary goal from form submission to the qualified stage, and keep form submission as an observation goal.
Which campaign brings qualified leads and which one only brings forms?
Ads Sensor reads every campaign by its goal type and switches the primary metric to the cost side for lead generation accounts.
Should you bid on volume or on quality?
With the same budget, bidding on volume buys more forms and bidding on quality buys fewer forms but more won deals. The deciding metric is not cost per form, it is cost per won deal. When you feed the qualified stage back, a drop in form count is the expected outcome and not bad news on its own. The number to watch is how many qualified leads enter the pipeline.
There is a cost to this switch: Smart Bidding needs volume to learn. If monthly qualified conversions per campaign fall into single digits, the system cannot separate signal from noise. In that case it is healthier to use a scored proxy value instead of targeting the qualified stage directly, or to consolidate campaigns so the data pools in one place.
How do targeting and form friction change quality?
Friction is the cheapest quality dial you have. Shortening a form lifts conversion rate but filters nobody; adding a qualifying question lowers volume but returns information sales will use in the first 48 hours. The rule is simple: only ask a question if the answer changes how the lead is handled. If it does not, that question is pure friction.
- Qualifying question: add one field sales genuinely uses, such as budget range, company size or intended use case.
- Multi-step form: put the easy questions first and contact details last. In practice this layout protects total conversion while making filtering easier.
- Confirmation step: accidental submissions are common on mobile placements. A review screen before submit noticeably cuts ghost records.
- Placement and match: broad automatic placements buy cheap forms. If your qualified rate is low, look at the placement and search term breakdown.
- Negative keywords: exclude low-intent phrases such as 'how to', 'internship' or 'template download' in search campaigns.
How do you build the feedback loop in practice?
The loop has four steps, and quality stops wherever the weakest link sits. No stored click identifier means no match; no sales scoring means no signal; a delayed upload means delayed learning. In practice the second link breaks most often: when the sales team does not mark records consistently in the CRM, the ad side keeps optimising blind.
Which metrics matter in long sales cycles?
If your sales cycle runs for weeks, this month's spend does not show up in this month's revenue. That is why lagging metrics need early indicators next to them: qualified rate, speed to first contact and stage-to-stage conversion. These reveal campaign quality long before closed-won data arrives.
- Qualified rate: qualified records divided by total forms. Track it by campaign, keyword and creative.
- Cost per qualified lead: spend divided by qualified count. This is the metric that should replace cost per form.
- Stage conversion: qualified to proposal and proposal to won, as percentages. Wherever the drop starts is where the problem is.
- Lag time: average days between the form and the qualified flag. This sets the ceiling on how fast bidding can learn.
- Win rate by channel: the same qualified count produces different win rates per channel. This ratio drives budget shifts.
The second half of the quality loop is reading the leads you already have. To see which segment genuinely comes back, read our guide to customer segmentation and RFM analysis, and to harden the signal layer underneath, see first-party data strategy. If you want to see which of your own campaigns brings qualified leads, you can apply for the beta.