Review your first 20 inquiries before judging the ad report
A short review of consecutive inquiries can reveal which records need checking. Keep contact events, opportunities and unverified outcomes separate.
TL;DR
- Choose a start time and review the next 20 inbound contact events. Keep repeat calls, spam and records you cannot classify in the file.
- Give each contact its own reference and connect repeats to the same job or estimate. Record whether the customer is new, existing or unknown.
- Check booked, sold, completed and paid in full separately, through a stated review date. A missing record stays unknown until someone checks it.
- This is a small record check. It cannot establish your campaign's conversion rate, return on spend or an agency's contribution.
Download the blank inquiry review workbook. It contains 20 input rows, status dropdowns, counts of confirmed and unresolved outcomes, and field definitions. The file uses local references; customer contact details can stay in the system you already use.
An agency report says leads increased. Your office says too many calls were repeats or work you do not offer. Before buying more traffic, put a short run of those inquiries next to the job records. You want to find the questions a summary has concealed: which contact belongs to which opportunity, what the shop could serve, and what has actually happened so far.
Google Ads reports conversion actions configured in the account. Its primary conversion column may include modeled results. Google's counting settings also operate per conversion action and ad interaction. A contact-event ledger has a different unit. You should expect to investigate the connection rather than assume its row count will equal a platform total.
The workbook is an original Lonta review method. Twenty is a manageable starting size we chose for this exercise, not a statistically justified sample size or an industry standard.
Set the scope before choosing the rows
Write down the source system, the start time and the business timezone. Then take the next 20 consecutive inbound contact events within that declared scope. A scope could be all shop inquiries or one recorded channel. If fewer than 20 have arrived, review the records available and state the count.
Include the contacts you would rather leave out: a repeated call, a message from outside the service area, an apparent spam form, or a record whose source is uncertain. Choosing only answered or successful inquiries changes the question before you begin.
Record the actual received time where available. Keep the source uncertain when the evidence is uncertain. A customer saying they found you on Google does not establish a particular paid campaign. If the selection was all shop inquiries, you cannot label the resulting file a Google Ads lead sample merely because some rows mention Google.
This review works with the records you already hold. It does not require new tracking software or uploading customer details to Lonta. The agency report guide explains the platform columns you may be comparing against.
Keep contact and opportunity references separate
A homeowner submits an estimate form, calls about it tomorrow and calls again to reschedule. Those are three contact events about one opportunity. Give each event its own inquiry reference and use the same opportunity reference for the related job or estimate.
In this worksheet, the first qualifying contact is labeled Eligible. Subsequent contacts about the same opportunity are Duplicate. A previous customer asking for a separate job can still be eligible; record Existing in the customer-type column. New-customer acquisition and work for an existing customer answer different business questions.
| Field | What to record |
|---|---|
| Inquiry ref and received time | The contact event you can find again |
| Source + evidence | A recorded channel or campaign and its local evidence reference; Unknown when unverified |
| Job type | The work requested, using the shop's terms |
| Customer type | New, Existing or Unknown after checking the customer record |
| Disposition | Eligible, Duplicate, Spam, Outside area, Service not offered or Unknown |
| Opportunity ref | The job or estimate reference shared by related contacts |
The workbook counts stage outcomes only for rows labeled Eligible. It trusts the labels you enter and does not automatically resolve matching references. Check repeated opportunity references before treating those counts as distinct jobs.
Keep a separate disposition for work outside your service area and work you do not offer. They require different questions about targeting and qualification. Keep Unknown for an unresolved record. It is a question to investigate, not a spam judgment.
Check outcomes as of a date
For each eligible record, use Yes, No or Unknown for each stage. Yes requires evidence. No means someone checked and the event was absent through the review date. Unknown means you have not established the answer. A blank status is included in the workbook's unresolved count.
| Stage | What a Yes means in this worksheet |
|---|---|
| Answered | A live answer or a person's response, beyond an automated acknowledgement |
| Booked | A service visit or estimate appointment is booked |
| Sold | The customer accepted paid work under the shop's documented rule |
| Completed | The work is documented as complete |
| Paid in full | The full payment is recorded; a deposit alone does not qualify |
These are separate observations. Booking an estimate does not establish sold work. Accepting work does not establish completion or collection of the full balance. Write the date through which you checked the records in Reviewed on.
A No recorded today can become a Yes later. For comparing different groups of leads, preserve their time to reach a decision. The lead conversion delay guide shows an equal-age comparison and explains why a partially observed group can mislead.
Twenty invented inquiries, four questions
The following is a constructed example with no client data. Its 20 contact events have these dispositions:
| Disposition in the invented file | Contact events |
|---|---|
| Eligible first inquiries | 8 |
| Repeat contacts about existing opportunities | 4 |
| Spam | 3 |
| Outside area | 2 |
| Service not offered | 1 |
| Unresolved | 2 |
| Total | 20 |
Six of the eight eligible records are from new customers; two are from existing customers. Through the invented review date, six eligible records have a confirmed booking, three have sold work, two have completed work and one is paid in full. One eligible record's stage outcomes remain unknown.
The invented example records make those counts inspectable. They are an illustration of the schema, not a benchmark to apply to your company.
This file suggests four checks. Link the four repeat contacts to their original opportunities. Check targeting for the two outside-area contacts. Check what the service-not-offered inquiry requested. Find the missing evidence for the two unresolved records and the eligible record whose outcomes are unknown.
It does not tell you to reduce spend. Eight eligible records in this deliberately constructed file do not establish a market-wide lead-quality rate. The remaining dispositions also do not prove a billing credit is due. Those decisions require the relevant platform rules, a complete matched population and the actual costs and outcomes.
What to ask your agency next
Bring the record references to the review. Ask which conversion actions feed the reported lead total and how repeated contacts or multiple actions are handled. Then ask how source evidence connects each eligible opportunity to a channel or campaign, which customer types are included, and what dates define the result.
If attribution or outcomes cannot be established, ask for the missing evidence and a person responsible for resolving it. You can improve the next report without pretending this short review has already measured advertising effectiveness.
For a broader review of your service area and measurement gaps, request Lonta's free strategy. You can also keep using this worksheet independently of hiring an agency.
What this doesn't cover
Twenty consecutive contacts are a practical diagnostic exercise, not a representative sample or a controlled experiment. The file contains no revenue, margin, media cost or management-fee calculation. It does not establish incremental demand, causal advertising effects, a channel ranking or a complete month's results.
Dates, source evidence and job status must come from your own records. The workbook summarizes its 20 designated rows only; start another copy for another review. Its counts are based on the labels you enter. They do not replace CRM reconciliation, validate customer identity or automatically deduplicate opportunities.
Sources and method
Google's conversion tracking definitions explain the configured actions and reporting dates. Its counting options describe One and Every per tracked action and interaction. Read October 13, 2026. The worksheet's definitions are our operating choices, not settings mandated by Google.
Lonta AI prepared the original worksheet, example and article with OpenAI Codex. The example is synthetic; no client observations or independent human review are claimed. The original worksheet and example are available under CC BY 4.0. Source-platform material retains its own rights.