A sourcing scorecard answers one question each quarter: for candidates who first inquired in the same period, which sources produced signed franchises, at what included cost, and how much of the picture is still open? Lead volume and Qualified rate are on it, but they don’t decide anything alone.
Most franchisors draw candidates from several places at once: paid social, paid search, franchise portals, brokers, referrals and existing franchisees expanding. This guide covers what each source needs from you, the scorecard columns and their denominators, and a fully worked fictional cohort you can check line by line.
The sources and what they ask of you
| Source | How candidates arrive | What you control | What to watch on the scorecard |
|---|---|---|---|
| Paid social | Ads shown to people who match an owner profile | Creative, audience, form, budget | High volume, lower Qualified share; rep hours per signing |
| Paid search | People already searching for a franchise in your category | Keywords, landing page, budget | Smaller volume, often higher Qualified share |
| Franchise portals | Listings on franchise directory sites | Listing copy, fee basis | Candidates often inquire with several brands; speed matters |
| Brokers | A broker introduces a matched candidate | Which brokers, the agreement terms | Large fee per signing; fewer rep hours; contractual credit |
| Referrals | Franchisees, staff, customers | A referral program, if you run one | Small and cheap; hard to scale on demand |
| Existing franchisees expanding | Multi-unit growth inside the system | Expansion criteria | Report separately: not new owners |
| Unknown | Records with no usable source | Your tracking | A row of its own, never spread across the others |
The scorecard columns
Every column has a stated denominator. Change any one of them between quarters and the comparison breaks.
| Column | Definition |
|---|---|
| Inquiries | Unique people whose first inquiry fell in the cohort period |
| Graded / Not yet graded | Graded = has a grade of Qualified, Nurture or Disqualified. Not yet graded is a status, counted separately |
| Qualified rate | Qualified divided by graded, not by all inquiries, so ungraded records don’t dilute it |
| Net signed within 180 days | Signed agreements, net of reversals, within 180 days of each candidate’s first inquiry |
| Signed per Qualified | Net signed agreements divided by the candidates ever confirmed Qualified in the observation window (first inquiry in the cohort period, first Qualified assessment within the same 90- or 180-day window after that person’s inquiry). That population is fixed once the window closes; it is not the count holding a Qualified grade at the cutoff, which can change as people are regraded. Count one agreement per signing candidate, so the ratio is also the share of those candidates who signed |
| Open late stage | Candidates past discovery day or in the disclosure period, unsigned at cutoff |
| Included cost | Media, management, production and tools, source fees (portal fees, broker success fees, referral bonuses) and a recruitment staff allocation: rep hours logged on that source’s records times one loaded hourly rate. Channel invoices are charged to the platform that billed them, even when some candidates that platform reached end up with an unknown source; any spend you can’t charge to a source stays on its own unallocated line so the total still reconciles |
| Included cost per signed franchise | Included cost divided by net signed. “Not calculable” when net signed is 0 |
| Rep hours per signed franchise | Rep hours divided by net signed |
| Median days to sign (signed only) | See the survivorship note below before using it |
Why staff time is in the cost. Leave it out and every high-volume source looks cheaper than it is, because volume is what eats rep hours. Allocating by hours logged, rather than by inquiry count, keeps brokers from absorbing teaching time they didn’t cause. House default is logged rep hours x the loaded rate. If logs are unavailable, a known recruitment staff-cost pool may be spread by graded-inquiry counts as a labelled fallback, using the same pool and denominator across cohorts and sources. That is an allocation convention, not measured time consumed; see the allocation policy and the fallback example in article 05.
The worked cohort (fictional)
Northfield Inspections is a fictional US home inspection franchisor. The cohort is everyone whose first inquiry fell between 1 January and 31 March 2026. The cutoff is 30 September 2026, so every candidate has had at least 183 days and the 180-day window is complete. USD. Staff time uses a $35/hour base assumption, with $50/hour as the high case. The 2026 source note explains the small salary sample, broad benefits loading and rounding. These rates are researched assumptions; all Northfield costs, hours and outcomes remain fictional.
| Source | Inquiries | Graded / not yet graded | Qualified (rate) | Net signed, 180 days | Signed per Qualified | Open late stage | Included cost | Included cost per signed franchise | Rep hours per signed |
|---|---|---|---|---|---|---|---|---|---|
| Meta (paid social) | 260 | 248 / 12 | 37 (14.9%) | 3 | 8.1% | 4 | $28,900 | $9,633.33 | 46.7 |
| Google Search | 70 | 68 / 2 | 18 (26.5%) | 2 | 11.1% | 2 | $13,325 | $6,662.50 | 22.5 |
| Franchise portal | 90 | 84 / 6 | 9 (10.7%) | 0 | 0.0% | 1 | $8,050 | not calculable | not calculable |
| Broker | 12 | 12 / 0 | 7 (58.3%) | 2 | 28.6% | 1 | $50,700 | $25,350 | 10.0 |
| Referral | 8 | 8 / 0 | 5 (62.5%) | 1 | 20.0% | 0 | $2,280 | $2,280 | 8.0 |
| Unknown source | 15 | 13 / 2 | 2 (15.4%) | 0 | 0.0% | 0 | $350 | not calculable | not calculable |
| Total | 455 | 433 / 22 | 78 (18.0%) | 8 | 10.3% | 8 | $103,605 | $12,950.63 | 34.1 |
Cost lines behind the included figures: Meta media $18,000, management $4,500, production $1,500, 140 rep hours ($4,900). Google media $9,000, management $2,250, tools $500, 45 hours ($1,575). Portal fees $6,300 plus 50 hours ($1,750). Broker success fees 2 x $25,000 plus 20 hours ($700). Referral bonus $2,000 plus 8 hours ($280). Unknown: 10 hours ($350) of logged rep time. That is the allocation policy, not a measurement of what unknown-source candidates cost to acquire: every channel invoice in this cohort is charged to the Meta, Google, portal, broker or referral row that billed it, even though some of the 15 unknown-source people may have been reached by one of those channels. Unallocated spend is $0 here, and the rows sum to the full $103,605. So the unknown row’s $350 is not a fully allocated cost, and it shouldn’t be set beside the other rows as if it were cheap.
In this cohort the 78 candidates ever confirmed Qualified are the same 78 who hold a Qualified grade at the cutoff, because in this fictional ledger all 78 were confirmed Qualified within their first 90 days and none was regraded before 30 September. Thus the 90- and 180-day Qualified populations happen to coincide here. Outside this example, count first qualification within the particular window being compared; a late qualification cannot be added retrospectively to its earlier window. All 8 signers were confirmed Qualified before signing, and each signed one agreement. If a grade had changed, the denominator would still be the ever-confirmed count.
What the cohort says:
- Volume and value disagree. Meta brought 57% of inquiries (260 of 455) and has the second-lowest Qualified rate. It still produced 3 of the 8 signings at $9,633.33 each.
- Brokers are expensive per signing and cheap in rep time. $25,350 per signed franchise, but 10 rep hours each against Meta’s 46.7. Whether that trade is worth it depends on territory and capacity. The brokers or ads decision works through it.
- Portals are “not calculable”, not zero. No signings in 180 days and one candidate still at a late stage. A $0 in that cell would say the portal was free.
- The unknown row stays. Fifteen people, two Qualified, no signings. Spreading them across the other rows would make every source look slightly better tracked than it is.
Staff-rate sensitivity
| Source | Included cost at $35/hour | Cost per signed at $35/hour | Included cost at $50/hour | Cost per signed at $50/hour |
|---|---|---|---|---|
| Meta (paid social) | $28,900 | $9,633.33 | $31,000 | $10,333.33 |
| Google Search (paid) | $13,325 | $6,662.50 | $14,000 | $7,000 |
| Franchise portal | $8,050 | not calculable | $8,800 | not calculable |
| Broker | $50,700 | $25,350 | $51,000 | $25,500 |
| Referral | $2,280 | $2,280 | $2,400 | $2,400 |
| Unknown source | $350 | not calculable | $500 | not calculable |
| TOTAL | $103,605 | $12,950.63 | $107,700 | $13,462.50 |
The direct-cost pool is $94,050 in both cases. Logged staff time totals 273 hours: $9,555 at $35/hour or $13,650 at $50/hour. Unknown-source staff allocation is $350 or $500 respectively; neither is a fully allocated acquisition cost. Signed counts, grading counts, rep hours and the source ordering are unchanged. Download the full base/high-case calculation.
The survivorship limit on time-to-sign medians
The tempting extra column is median days from inquiry to signing. For this cohort it would read: broker 74 days (61 and 87), referral 61 (one case), Google 96 (88 and 104), Meta 118 (102, 118 and 141), all sources 95.
Those numbers describe only the candidates who had signed by the cutoff. Everyone else is censored: the 8 candidates still at a late stage, the other Qualified and Nurture candidates who haven’t signed, and anyone whose eventual outcome (signing in month eight, signing with another brand, or never signing) isn’t known yet. A median taken over signers alone can’t tell you what the eventual median would be once those outcomes arrive, or which direction it would move, and the effect is largest for sources with many open cases, such as Meta with 4 open late-stage candidates against 3 signings. With 1 to 3 cases per source, a single signing moves the median by weeks. The observed figures above are exact for the signers they describe; that’s all they are.
Use share signed within a fixed window (90 or 180 days, measured over all candidates ever confirmed Qualified in the window) as the comparable speed measure. It counts every candidate in the denominator, signed or not. Keep the signed-only median as a footnote with its case count, or leave it off.
Turning the scorecard into a mix decision
- Wait for the window. Compare sources only on cohorts whose window is complete for every candidate. Northfield’s Q2 cohort closes its 180-day window on 27 December 2026.
- Repeat before you move large budget. One cohort with 1 to 3 signings per source is an observation. Further cohorts pointing the same way add context and make a decision easier to defend, but no fixed number of cohorts turns small counts into proof.
- Read capacity next to cost. Northfield’s paid sources need 22.5 to 46.7 rep hours per signing, against 10 for the broker. Adding paid volume without adding hours produces Not yet graded records, not signings.
- Change one source at a time where you can, so the next cohort can tell you something.
Check the page before you blame the source
Every source on the scorecard sends candidates to a page and a form. If the page doesn’t state the money, candidates who can’t meet it can inquire from any source, and a low Qualified rate can reflect the page rather than the channel.
In our audit of the top 100 franchise brands’ recruitment pages (checked 31 October 2026), 30 of the 69 pages we could score showed neither an investment figure nor a capital requirement. 31 had a readable inquiry form; 16 of those forms asked about capital, 13 about territory and 1 about timing.
Before you cut a source for a low Qualified rate, check that its candidates landed on a page that shows the investment and a form that asks capital, territory and timing. The inquiry form spec has the wording.
Start here
Pick the most recent inquiry quarter that is at least six months old. Tag every inquiry with its original source, add an unknown row, and fill in the columns above with rep hours included. If the unknown row is big enough to change which source ranks first, fix source capture before you trust the rest of the table.
Download: S11-sourcing-scorecard.csv (the worked cohort with every cost line) and S11-signing-log.csv (each fictional agreement with inquiry and signing dates, so you can recompute the windows and medians).
Related: Brokers or your own ads · Cost per signed franchise · Source tracking · Recruitment report
Put it to work on your recruitment system
Bring your recruitment page and anonymized stage counts to a diagnostic. We’ll identify what to test and how to measure it.
Discuss your franchisor recruitment system