The First Question I Ask When A Contractor Says The Leads Are Too Expensive

Why the vendor is not the first question

The vendor, campaign, price, and source settings may matter. They simply arrive after the first evidence question. A paid inquiry can be wrong for the service area. It can also be a reasonable fit that entered an unowned queue, received only an autoresponder, lacked a second touch, or stopped after an estimate because the deposit or next step was not explained in the record. Those are different situations and should not be compressed into one complaint.

The last question creates a reviewable path. It does not excuse poor targeting or expensive media. It prevents the source decision from outrunning the available record.

The first fields to inspect

Write down the responsible role even when the inquiry was a poor fit. For example, the office may have replied and confirmed that a requested service falls outside its area. That record belongs in the fit discussion, not in an unanswered-inquiry total. Keep the sample's limits visible: a handful of selected records cannot establish a campaign-wide rate.

Use a redacted mixed sample

A first review does not need full CRM access, customer lists, private messages, complete contracts, login credentials, or payment records. A small redacted sample can compare different handoff states: one supported fit with a next action, one wrong-area or wrong-service inquiry, one duplicate, one missing owner, one missing response, one estimate with no supported deposit or next-step note, and one source label that remains unknown.

That sample is illustrative, not a customer case and not a required count. Redaction does not guarantee anonymity; uncommon jobs or combinations of details can still identify someone. Remove names, phone numbers, emails, exact addresses, private notes, pricing details not required for the question, credentials, tokens, and unrelated history. A row should contain enough role-level evidence to classify what is visible while leaving unsupported conclusions unknown.

A good record may still end in no booking. A weak record may coincide with a sale. The review grades evidence quality, not customer value or employee performance.

Separate fit from follow-up

Mixing wrong-fit leads with unanswered good-fit leads makes the phrase “too expensive” harder to inspect. Separate at least four evidence buckets:

• Supported wrong fit or duplicate: area, service, timing, or repeated-inquiry evidence explains the label. Track duplicates separately from unique inquiries when comparing counts.
• Follow-up evidence to check: an owner, response, last note, or next action is absent from the reviewed record. That absence alone does not prove that nobody acted.
• Supported handoff: the record shows a responsible role, useful response, fit decision, and next step.
• Unknown: the sample lacks enough evidence to classify the path.

Unknown is not a failure label. It is an instruction to identify the smallest missing proof: a callback note, source mapping, service-area rule, estimate step, portal event, or route test.

Read a missing note carefully

Consider two fictional records selected to illustrate the distinction. Record A identifies the requested service, an office owner and a call time, but the notes export contains no reply entry. Record B includes a dated staff note saying the caller was outside the service area, along with the recorded response. Neither record includes a verified final job outcome.

For A, write "response not visible in this export" and ask the responsible role which system holds the callback record. A separate phone log might show an attempt; it might show only a routing event. Until that check is made, "the office never called" goes beyond the evidence. For B, retain the documented area mismatch and response. Do not relabel it as an unanswered lead because it did not become a job.

These invented rows demonstrate how to label evidence; they are not customer cases, a recommended sample size or a comparison of vendors. They also cannot produce a cost-per-job figure. That calculation needs a consistent time period, verified spending and a defensible job count. Keep the calculation separate until those inputs are available, and record any duplicate or credit treatment so another reviewer can reproduce it.

Build a simple lead-cost cleanup board

The smallest proof point should match the suspected handoff. For missed calls, use a redacted call log and callback note. For paid leads, use a small source, owner, and first-response sample. For stalled estimates, use the sent date, deposit or next-step note, and last supported status. For a messy booking handoff, use the booking request, role-level owner, and supported status note. These are evidence requests, not findings.

This is a cleanup surface, not a revenue forecast or vendor scorecard. It shows where the next inspection belongs without converting missing data into source quality, buyer intent, staff fault, savings, bookings, sales, or revenue.

Resist fast conclusions

The FTC's Advertising and Marketing Basics and Policy Statement Regarding Advertising Substantiation address support for advertising claims. They do not prescribe an internal contractor lead review. The narrower lesson used here is that strong claims need appropriate support; it is not an FTC endorsement of this method.

A six-step lead-cost review

1. Define the cost concern. State which source, period, lead type, and business decision are actually under review.
2. Ask what happened after arrival. Start with the handoff record instead of assuming the vendor, office, or buyer caused the problem.
3. Build a small redacted sample. Keep source, generalized job type, owner, first response, fit, last note, and next action while removing customer identity.
4. Separate fit from follow-up. Group wrong-area, wrong-service, duplicate, unanswered, and supported next-action records without assigning unsupported blame.
5. Name the missing evidence. Record unknown owner, response, fit, note, route, timing, or source data instead of forcing a conclusion.
6. Choose one bounded next decision. Select a route check, response cleanup, label fix, vendor review, volume hold, or no-change decision supported by the sample.

When a source pause may deserve review

Pausing or reducing a source may be reasonable to evaluate when supported records show persistent wrong-area or wrong-service mismatch, documented duplicates, settings that conflict with the actual service area, operational volume beyond safe capacity, more operational confusion than the business can safely handle, or a cost concern that remains after ownership and response evidence are clean. The decision still needs verified spend, contract, credit, timing, and campaign context.

A clean record cannot promise savings, refunds, improved vendor behavior, better lead quality, lower future costs, recovered work, sales, or revenue. It can make the next decision more specific and reduce the number of unknowns being treated as facts.

Method and source basis

Bring the field list to the person who handles incoming inquiries and ask where each missing event would be recorded. Use the blank board for the answer, including "not available" when that is all the team can establish. If you want an outside review, read the scope on the order page first. A review of these records does not include an automatic vendor rating or a revenue forecast.

The first-proof-point request should stay narrow: a redacted source and handoff sample, generalized call states, estimate timeline, route label, or role-level status. The related note-check article covers the “nobody replied” claim, while the revenue leak calculator is an illustrative planning tool, not evidence of a specific loss.

No source verifies a contractor, vendor, campaign, lead, cost, source quality, staff action, buyer intent, refund, saving, recovered job, sale, revenue effect, ranking, traffic, or AI citation. The field map, evidence buckets, and six-step review are AI Cleanup Doctor operating guidance and require human judgment.

Frequently asked questions