Define a lead your team can actually work
A form submission is an inquiry. It becomes useful when you know who is asking, what they need, whether your business can serve them and what happens next. Agree on a small set of qualification rules with sales before changing your campaigns.
Your definitions should fit your business. A hotel group inquiry may need dates, room count and availability. An automotive customer may need fitment confirmed. A service business may need location, job type and scheduling capacity. Do not make customers complete an interrogation to ask a reasonable question. Collect enough to route the conversation, then qualify it properly.
Measure the handoffs, not just the totals
Give every inquiry a consistent status and an owner. Keep duplicate contacts attached to the same opportunity. Then track where people move forward, where they stop and why. Show the underlying counts beside each percentage. A rate on its own can conceal a very small sample.
| Measure | What I want it to reveal |
|---|---|
| Contact rate: reached ÷ valid inquiries | Can your team start a real conversation? |
| Booking rate: booked ÷ qualified opportunities | Is the next step clear and available? |
| Show rate: held ÷ scheduled appointments due | Are expectations and reminders working? |
| Close rate: wins ÷ completed sales opportunities | Does the offer solve the customer's problem? |
| Collected revenue and contribution | What did the business actually receive and retain? |
Find the economics behind a cheap lead
Here is a hypothetical comparison. One source delivers 100 leads at $50 each and produces five new customers: $1,000 in media cost per customer. Another delivers 40 leads at $100 each and produces eight customers: $500 per customer. The second source has a higher cost per lead and a lower media acquisition cost.
Neither result is fully loaded CAC until you add the other acquisition costs. Neither proves profitability until you review margin, cancellations and collections. Use the lead funnel calculator to change the assumptions and see which handoff has the greatest practical effect. Capacity matters: there is little value in buying appointments your team cannot serve.
Teach your advertising what a useful outcome looks like
Google Ads supports qualified-lead and converted-lead goals informed by offline outcomes. A converted lead represents the stage your business defines; it does not automatically mean cash collected. [1] Your team needs a shared definition before an integration can send a meaningful signal.
Audit the actions your campaigns treat as success. A page visit, short call and completed sale should not quietly become interchangeable. Keep diagnostic actions visible, but have your specialist review which outcomes guide bidding, their reliability and their reporting delay. Recent campaigns may look weaker while conversions are still developing. [2]
Use AI to improve the follow-through
I would start with controlled tasks: drafting a response from approved information, summarizing an authorized customer conversation, or flagging opportunities missing a next step. Your team should verify the facts and the recommended action before anything sensitive is sent or changed.
For an AI booking assistant, test the difficult cases: unavailable appointments, duplicate inquiries, incorrect locations, cancellations and requests for a person. Review completed appointments and correction time alongside response speed. As you plan for 2027, build a process you can audit. Faster answers help when they are accurate, useful and connected to someone who can deliver.
Sources & context
Checked October 7, 2026. I use these sources for the facts and definitions noted above. The questions and suggested actions reflect my perspective; adapt them to your business.
- Google Ads: Qualified leads and converted leads
Defines the two goals and the role of offline lead outcomes. Businesses define the conversion stage.
- Google Ads: Conversion lag reporting
Explains the delay between an ad interaction and recorded conversion. The funnel definitions and examples above are operating recommendations, not universal benchmarks.