How to audit your lead response before buying more traffic

Do you need more leads or better follow-up?

More traffic deserves a closer look when suitable inquiries already receive a useful response and a clear next step, but there are too few of them. Follow-up deserves attention when inquiries arrive and then sit unassigned, calls go unreturned, or nobody can explain what happened after an estimate.

Both problems can exist at once. The purpose of an audit is to separate them before choosing a bigger ad budget, a new website, or an automation. Lead volume alone cannot tell you which investment to make.

Choose a sample you can describe

Start with a fixed period, such as the last two complete weeks, and include inquiries from calls, forms, email, and other channels you actively manage. If volume is low, extend the period and record the dates. If volume is high, define a sampling rule before looking at the outcomes.

Include missed calls and unanswered forms. Reviewing only the inquiries that became customer records can hide the ones that never made it that far.

Keep customer details in your approved business systems and use non-identifying inquiry IDs for a shared review. The review should make missing ownership visible without creating another customer database.

Trace six points for every inquiry

1. Arrival. What channel received it, and at what time? Preserve the original timestamp and time zone.
2. First useful response. When did someone answer the request or establish a clear next step? Log an automatic acknowledgment separately.
3. Owner. Who was responsible for moving it forward? A shared inbox is a location, not an owner.
4. Next action. Was a callback, visit, estimate, or clarification scheduled? Record the due time.
5. Latest outcome. Is the inquiry open, booked, quoted, won, lost, duplicate, spam, or outside scope?
6. Evidence. Which call record, message, calendar entry, or job record supports the outcome?

Keep inquiries, estimates, signed jobs, completed work, and collected revenue distinct. They answer different business questions. An estimate sent is not a job won.

Calculate four useful measures

Define “eligible” before comparing periods. For example, you might exclude confirmed spam and duplicates while keeping unanswered requests whose fit is still unknown. Record every exclusion and its reason. Do not quietly remove difficult inquiries to improve the result.

Use elapsed time as the default. If you also calculate business-hours response time, show it separately and state the opening hours used. For a small sample, show the actual counts next to the percentages.

Read a small sample without overclaiming

Illustrative example: a team reviews 30 inquiries received during a defined period. Five are confirmed duplicates or spam, leaving 25 eligible inquiries. Four have no useful response by the audit cutoff. The unanswered share is 4 ÷ 25, or 16%.

That finding identifies a follow-up problem to investigate. It does not tell us that four jobs were lost, that AI would have won them, or what revenue was missed. Those conclusions need additional evidence. These figures are invented to explain the calculation, not client results.

If six of the 25 inquiries are still waiting for the customer to choose an appointment, the booking share may change. Record the observation window and revisit the same cohort rather than presenting its early result as final.

Fix the handoff with the clearest failure

Group exceptions by the step where the inquiry stopped. A broken form notification calls for a delivery fix. A callback without an owner calls for assignment. A promise of an unavailable appointment calls for a calendar or instruction fix.

• Missing arrival: run a controlled test through the form or phone number and check the destination.
• No owner: assign one person and define a backup for absences.
• No next step: agree on what the customer and team should expect next.
• Repeated unanswered calls: review staffing, routing, overflow, and the callback process before selecting a tool.
• Conflicting records: choose the system of record and reconcile duplicates.

Change one part of the process, then review a comparable period using the same definitions. If traffic sources, staffing, or seasonality changed too, record that context. A before-and-after difference alone does not prove the fix caused it.

Use the findings to choose what to build

The audit should end with a short action list: the failure, the person responsible, the proposed fix, the due date, and how you will verify it. Some fixes are administrative. Others may need a better website, phone setup, or connection between tools.

If you identify a repeated language task, use Where to start with AI in a service business to scope a pilot. If call answering is the issue, use the AI receptionist evaluation checklist before moving customer calls.

Our Bakeris case study describes work connecting inquiries, office follow-up, and reporting. For OneCompass’s broader implementation work, see our services. The same principle applies across these projects: make the next step clear, give it an owner, and keep the outcome visible.