Franchise lead generation for franchisors

Franchise lead generation, for a franchisor, means bringing prospective franchisees into a development process your team can work. When the pipeline feels thin, the first question isn't which channel to add. It's where the inquiries you already paid for went. Trace one month of inquiries from form submission to first conversation, count what dropped out at each step and why, and fix the largest leak you control before you raise the budget.

This guide shows that trace on one fictional month, with every count reconciled, and the decision it supports.

What "lead" means at each step

A form submission is an event. A unique inquirer is a person. A Qualified candidate is a person who meets your written standard. A first conversation is a conversation that happened, not a booking. Each is a different denominator, so "cost per lead" can mean four different numbers.

Grading follows the shared rules in our qualification guide:

• Qualified (green): every must-have rule passes explicitly.
• Nurture (amber): suitable, with a dependency your rep has assessed and dated: a funding plan with a named source and expected date, a named open territory the candidate accepted while their first choice is reserved, or a start date 6 to 18 months out. A missing answer isn't a dependency.
• Disqualified (red): fails a documented must-have, with the reason recorded.
• Not yet graded is a status, not a fourth grade. It covers missing answers and records waiting for a rep's confirmation, and is counted separately so a backlog never looks like poor quality.

Grade is about fit. Whether someone answered the phone is lifecycle, recorded separately. A Qualified candidate who hasn't responded is still Qualified.

The trace: one month, every inquiry accounted for

Fictional example. A US franchisor, USD, using the shared rules plus one rule of its own (R6: wants a job, not a business). Inquiries received 1 to 30 September 2026. Status read on 21 October 2026, the cutoff this team set to fit its two-business-day grading target. Media spend in September: $9,300 across Meta and Google search. These are not SMF results or industry benchmarks.

Someone who hasn't replied isn't invalid. Only contact details shown to be wrong are removed here; non-response stays in the valid count and is tracked as lifecycle.

The 156 graded inquiries:

Check: 31 + 58 + 67 = 156; 21 + 9 + 28 = 58; 24 + 19 + 15 + 9 = 67.

The 31 Qualified candidates, by lifecycle:

The first three rows sum to 31, and so do the last three. They're two views of the same people. Net signed agreements from this cohort at the cutoff: 0.

The same spend, four different costs

These are blended media indicators: all September spend divided by all of the cohort's counts, whatever the source. The last two say more about whether the pipeline can produce agreements. Cost per signed franchise isn't calculable at this cutoff because the cohort has no signings yet. The $9,300 was still spent, and the figure becomes calculable when a signing is recorded. Our cost per signed franchise guide covers cohort cutoffs.

Where to look first

Rank by people affected and whether the fix is yours. Then ask why, before deciding what the cause is:

1. 3 Qualified candidates never contacted, 6 contacted late. Small numbers, but these are the candidates the budget exists to find, and the fix is assignment, not spend. See how fast to contact a franchise candidate.
2. 21 not yet graded (11.9% of valid inquiries). Some may be Qualified. Clearing the manual-review queue costs staff time, not media.
3. 58 Nurture candidates. Each has a dated dependency. The leak is any record without an owner or a next date.
4. 19 who wanted a job and 15 who only wanted a closed territory. These are recorded reasons, not proof that the ads failed. Pull a sample of each and check the ad, placement and page they came from. A territory can also close after someone inquires, so check the territory status on the inquiry date before blaming targeting.

Split by first observed source

First observed source is where a record was first seen arriving. It isn't causal credit, and it isn't the platform's own attribution. The Unknown row may include people who came from a paid ad without the source being captured. That's why the per-source costs divide each channel's spend only by the candidates recorded against it, and the blended figure stays visible.

17 of the 19 job seekers had Meta as their first observed source: 13.2% of Meta's valid inquiries, against 4.9% for search. That's a reason to read the Meta ads and page through a job seeker's eyes, not a rule about either platform. The fix in this example is a wording test, judged by the next month's grades.

The decision this trace supports

For this fictional franchisor, the trace argues against raising spend this month:

• This week: assign and call the 3 Qualified candidates never contacted, clear the 21 records in manual review, and give every Nurture record an owner and a date.
• Next creative round: test Meta ad and page wording that names the owner role and lists open territories. Judge it by Qualified candidates and job-seeker disqualifications per week, not form volume. For writing the ads, see recruitment ad creative.
• Targeting: check location targeting and the form's territory list against the current territory status. Our territory availability guide shows one change across every surface.
• Re-run the trace on October with the same cutoff rule. A before-and-after comparison is an observation. The months differ in more than your edits.

If the trace showed an empty review queue, quick follow-up and few wrong-fit inquiries, but still too few Qualified candidates, then more budget or a new source would be the next question.

Run the trace on your own month

1. Pick a closed month and a cutoff you can defend from your own grading and contact targets. Write the cutoff on the report.
2. Export every delivered submission with: inquiry ID, contact ID, received time, first observed source, campaign and ad IDs, removal reason and evidence (for example the bounce record), grade, grade reason, rules version, first attempt time, conversation status and signed status.
3. Remove repeats by contact, then demonstrably invalid records, logging the evidence for each.
4. Count Not yet graded separately.
5. Fill the steps table, then check every split against its parent total.
6. Rank what you find by people affected and whether the fix is yours.

If a field doesn't exist, record that step as "not measurable" rather than estimating it. That gap is your first fix. Our ad leads versus CRM guide covers reconciling platform counts with CRM records, and the source tracking guide covers the source fields.

Download the worked month: C01-lead-leak-audit.csv (every step, reason, owner and fix) and C01-source-split.csv.