Lead Pronto — what it costs, what it books, and who they send

The last two columns are the mix: what share of that week's leads were under 35, and what share were 45 or over. Watch them move together with the booking rate — that is the mechanism, and it is why the CPL rising did not make the leads worse.

Weeks run Monday to Sunday, so the first and last weeks of the page are part-weeks. Any week with no CPL reported is outside the costed window: Lead Pronto report per month and Sep 2026 has not been reported yet. Those weeks show volume and bookings only — no cost is imputed for them anywhere on this page.

Ignore the booking rate on the last week or two. A lead has to be distributed before it can book, so the most recent weeks have barely started — their rate rises for days afterwards. The last full week on this table is the newest one worth reading.

Newest lead on this build: 2026-09-17 — the feed is current to today, so only the calendar makes the final week short.

The booking rate went from 4.42% in Jun 2026 to 5.20% in Aug 2026, +0.78pp. Two different things could have caused that, and they have very different consequences, so they are separated multiplicatively: hold one factor at Jun 2026 and move the other.

69% of the improvement is who they sent, not how the leads performed. That matters for what you do next: a mix change is something Lead Pronto can be asked to keep doing (or do more of), and it can be reversed next month without anybody breaking anything. A within-band improvement would have been a quality change and would have been more durable.

The two single-factor rows deliberately do not sum to the whole move. The leftover +0.14pp is the interaction — bands whose share grew and whose own rate moved — and it cannot honestly be handed to either factor. Folding it into one would overstate that factor by 17% of the whole move.

The grey number after each rate is that band's lead count in that month — the rate is only worth reading where the count is large.

The CPL column is the point. Lead Pronto bill one price per city per month, so the price a band carries here is only its city and month mix — and it barely moves ($16.26 to $17.41 across every band). Cost per booking moves by a factor of 3.9. Every bit of that is booking rate — and then the show rate makes it worse again.

Show rate is of the people who reached the calendar, how many turned up: (roster − no-show/cancelled) ÷ roster. Its denominator is the roster feed, not the booking count — those are two independent CRM reports that do not nest, which is why the 45+ bands carry more calendar rows than bookings. Under 40 calendar rows it is shown as a plain count, because a show rate on ten people is not a rate.

Revenue per $1 spent is the gross package value those leads signed divided by what they cost to buy. It is gross, before a single studio cost and before the collections book fails to collect part of it — so 1.00× is not break-even, it is a long way behind.

The show rate compounds the booking rate, it does not offset it. The bands that book worst also turn up least — 18-24 shows at 60.8% against 43.8% for 55+. So a band booking at a third of the average puts a person in a chair at less than a third.

Read the 55+ row twice. 3,959 leads (20.7% of everything Lead Pronto sent), $64,387 spent, and $64,252 of gross package revenue behind it — 1.00× the ad spend, against 4.13× for 18-24. Per lead bought that is $16.23 against $71.95. Every one of those leads cost the same money.

This is the clean comparison. Every lead in a row cost the same, so the colour is pure performance: green beats that city's own average, red is under 60% of it. The grey number is the lead count.

Yes — but almost all of the move is older than a month, and the last month has mostly held it. Against the previous 30 days the average lead is 1.1 years younger and the under-35 share is +3.0pp. Against the 30 days before that it is 4.9 years younger and +15.8pp. The step happened in early August; since then it has settled.

These are rolling 30-day windows ending today, not calendar months — a calendar month would have hidden a change that started mid-month, which is exactly what this one did. They run into the current month, which Lead Pronto have not costed, so there is no money anywhere in this section: mix needs no CPL.

A five-day-old lead has not produced less revenue than a sixty-day-old one. It has produced it less far. A lead has to be distributed, then booked, then turn up for a shoot that may be weeks out, then buy. Lined up side by side, the newest window will always look like a collapse, and it is not one — so read the revenue column as so far, never as a level.

The show rate holds up because it is a rate among people who already had an appointment — its numerator and denominator mature together. The revenue columns do not have that property, which is what the next table is for.

Every window gets the same 14 days of runway: only leads that have actually had 14 days are counted, and only bookings that happened inside them. Exact, not modelled — the bookings feed carries the booking date. This is the comparison the raw revenue column cannot make.

On a like-for-like 14 days, the newest leads book at 4.75% against 4.68% and 3.85%. So the drop in the raw revenue column above is maturity, not performance — the newest cohort is the strongest of the three at the same age.

Green is the direction that helps: more of the two bands that book, less of the two that do not. Anything under a point is drift, not a change.

This is why the monthly view is not enough. The mix did not drift — it stepped, in one week, and the average age has been creeping back up since that peak. Watch the Average age column: if it keeps climbing, the improvement is unwinding and it is worth asking Lead Pronto why before it has fully gone.

The CPL column is there for weighting, not comparison: it is what that city pays per lead across the whole costed window, so a city that is getting older at a high price is the expensive version of this problem. Cities with under 100 leads in either window are left out.

The last four column pairs are before-and-after, and only the first two pairs are safe to compare. Booking rate and show rate settle within days; revenue does not — a lead from last week has not had its shoot yet, so the "last 30" revenue column is unfinished in every single row and will keep rising for weeks. It is shown because you asked for it, in grey, as a running total — never as a fall.

Chicago moved +18.3pp younger while Boston moved -16.7pp older — 5.7 years older on the average lead in a single month. Those two cancel in the network number, which is the whole reason this table exists. Boston is the one to raise: it pays $25.13 a lead.

Two cities buying at the same price can still get very different books. This is where the order needs changing, city by city.

All 19,161 fresh Lead Pronto leads in this window are Female. Not most — all of them. Whatever is being asked of the gender axis, Lead Pronto cannot answer it: there is nothing to compare against.

This is a fact about the whole book, not about this window. Across the same three months the only supplier sending men at all is Alan. Blue Rooms and the in-house codes are effectively all-female too. So a question like "are male leads cheaper or better?" has to be put to Alan's book, on https://reports.iconicbyai.com/worst-leads or https://reports.iconicbyai.com/alan-leads — not to this one.

The useful version of the question for Lead Pronto is the one on the Age tab: they send one gender at every age, and the age is what decides the cost.

Source: the CRM's own Sex field on the lead-in feed, exactly as the lead arrived. Nothing is inferred from the name or the campaign code.

Straight-line miles from the lead's own 5-digit ZIP centroid to that city's studio — not the target sheet's per-prefix number, which only exists for prefixes that are already on a list and so reported every off-list lead as having no distance at all. 99.0% of the book is geocoded. Florida is measured to the nearer of Parkland and Orlando, because #FL shoots at both.

Half of all Lead Pronto leads live within 29 miles of the studio, and 66.0% of the book is inside 50 miles. The interesting half of the table is the right-hand end: 11.9% sits past 150 miles, and that share is wildly uneven by city.

Read a row left to right and it draws the city's shape. "Half are within" is the typical lead; "3 in 4 are within" is where the furthest quarter starts. When those two numbers are close together the city is one tight blob — Chicago is 14 and 34 miles. When they jump, the city has a second cluster sitting a long way out: Dallas goes 41 → 164 miles, and Houston 24 → 160.

The "Furthest lead" column is one single lead, not a pattern — every city has a handful of people who filled the form in from another state. Judge a city on the middle columns and on "Past 150 mi", never on its furthest.

Count and, in white, that band's share of the city's own book. The three far bands are coloured red throughout so the shape of each row reads at a glance.

Two cities are not shaped like the others. Dallas and Houston each carry roughly a quarter to a third of their book in the 151–250 mile band while every other city is near zero there. That is not a long tail — it is a second cluster, and the tail table at the bottom of this tab names which metro each one is actually buying.

Coloured against each city's own average, which is the only fair test — every lead in a row cost that city the same price. Cells under 100 leads show the count instead of a rate, because a rate on 40 leads is noise. The grey number after each rate is the lead count.

Note the near end: the closest band is not the best one. The 26–50 mile band out-books 0–15 miles, which rules out "closer is always better" as the rule to buy on. The break is further out, and it is sharp.

Partly — and the two tables below are how much. The far bands really do skew older, so some of the distance penalty is the age penalty already on the Age tab. But not all of it: in the rate grid, 18-24s still book worse the further out they live, which age alone cannot explain. Distance and age are separate charges on the same lead.

And the booking rate in each cell of that grid — cells under 60 leads show the count instead of a rate:

150 miles across Texas is not 150 miles in New England. This table uses Neil's own 3-digit-prefix drive-time sheet (drive_time_prefixes.json) instead of the crow-flies number. A prefix that is not on the sheet is off the studio's own map entirely and is shown as its own row rather than folded into a band.

The drive sheet has broken rows, and they are quarantined rather than used. Straight-line miles are the shortest any route can physically be, so a prefix whose recorded hours imply more than 75 mph in a straight line is a bad row on the sheet, not a fast road. The test also has to be material or it fires on rows that are merely coarse — plenty of same-metro prefixes are recorded as 0 hours, and a 26-mile prefix at zero still belongs in "up to 1h". So a row is only pulled when it is impossible and further than 60 miles, which is where the error starts moving a lead between bands.

1,356 Lead Pronto leads (7.1% of the book) sit on 48 such prefixes. Left in, most would have landed in "up to 1h" — a 365-mile lead counted as a local one.

This is a note for the sheet, not for Lead Pronto — /distance-age reads the same file and inherits the same rows.

2,256 leads (11.8% of the book) sit past 150 miles and cost $38,682. They book at 3.28% against 4.74% overall.

Read the City column against the State and ZIP3: this is where one city's campaign is buying another city's metro. Ordered by volume.

"Off list" means the lead's 3-digit ZIP is on no studio's target sheet (→ reference_ooa_ziptarget_rule) — a breach, not a distance judgement. Florida is scored against Miami and Orlando, the only two-studio market.

On its own tab — the matrix, the per-city distribution, drive time, and the named far tail all live under Distance.

A prefix full of retirees would book badly whatever the ZIP did, so every prefix is also scored age-standardised: its own booking rate inside each age band, re-weighted to its city's age mix. If the standardised rate lands close to the raw one, the age mix was not the cause and the ZIP is an independent problem — cutting an age band will not fix it.

Ordered worst cost-per-booking first. Every prefix with 120+ leads is listed; the full set including the small ones is in the CSV.

Worst cost-per-booking first. The colour on Books at is against that city's own average, which is the only fair test — every lead in a row paid the same price, and the city column next to it says what that average was.

Because the cost is flat inside a code-month, these are not projections: they are the same 19,161 leads re-totalled with a slice removed. The saving and the lost bookings both already happened. A cut is never free — the lost-bookings column is the number to weigh the saving against.

Lead Pronto's own under-35 leads cost $17.26 and book at 7.09%. The $64,387 now going into 55+ would, at that price, buy 3,731 under-35 leads and about 264 bookings, against the 72 the 55+ spend produced.

Two things in that sentence are not evidence, and must not be quoted as if they were. First, whether Lead Pronto can deliver that volume of under-35s at that CPL is UNKNOWN — we have never asked them to, the younger end is the harder end to source, and their price would be expected to rise as the target narrows. Second, the 7.09% is measured on the under-35s they send today; buying three times as many would reach deeper into the same audience and should be assumed to convert worse, not the same.

The defensible half of it is the half that already happened: $64,387 bought 72 bookings. That is a count, and nothing selected it.

Straight from Lauren Renshaw's (and Louis Melia's) own emails, converted at 1.3198 USD per GBP. Nothing here is our estimate of their price — it is their price. A cell marked part month is a month where only some weeks were ever reported, so the CPL is that period's, not the month's.

No report exists for the current month. Lead Pronto report per month and Sep 2026 has not been sent. Its leads appear on the trend tab with volume and bookings only; no cost is imputed for them anywhere.

Some months are built from weeklies, not a monthly. Where that is the case the code-month is flagged part month above and its CPL covers only the weeks that were reported — the spend it implies for the whole month is correspondingly uncertain.

They quote CPL, never spend. Every ad-spend figure on this page is derived: leads × their code-month CPL. The one real spend table they have ever sent was a May 2026 monthly summary. If a total here needs to be defended to the penny, it needs an invoice from them.

Where the emails live: leads@fivechickens.com on Outlook Web, not Gmail, and the subject line changes constantly — never search by subject, search by sender address (→ reference_lead_pronto_weekly_volumes).

Neil, 16 Sep 2026: "i think we have to allow the same lead cost for each bucket regardless." That is exactly how Lead Pronto sell. They bill one CPL per #{CITY}LP2 code per month and do not price a 55-year-old differently from a 22-year-old, or a Miami ZIP differently from an Orlando one.

So every lead inside a code-month carries the same cost here, and a bucket's ad spend is simply its lead count × that code-month's CPL. The consequence is the whole page: inside a city, a bucket is expensive only because it books worse.

Across cities the price does move — Boston runs about $25.13 a lead against New York's $10.08 — so every cross-city cut on this page prints its blended CPL beside it, and the Age tab carries a per-city matrix so the comparison can be made inside one price.

19,161 fresh Lead Pronto leads, 2026-06-01 to 2026-08-31 — the last three complete calendar months. Complete is not a style choice: Lead Pronto report per month, and a part-month at the end has no report to cost it.

The *LP2R recycled codes are out of every cost figure (651 leads in the window, 21 bookings). They were never bought a second time, so giving them a CPL would flatter the whole book. Dupes, removed and rejected leads all stay in: the question is what the money bought, and a lead the team had to throw away is exactly what makes a bucket expensive.

On name + marketing code, never name + date — Date Distributed is not Date Lead In. The most recent matching lead row takes the credit and each booking row is consumed once. Across the whole feed 8,278 of 8,810 booking rows matched a lead (94.0%); the rest belong to other suppliers, recycled rows and leads from before the window.

Booking, show, no-show and sale feeds are all read past the window end so a late outcome still lands on the lead that earned it. The flip side is that the newest month is the least mature figure on the page — a lead from its last week can still book after this page was built.

1. "Spend ÷ revenue" is gross package value, not cash. It is the ad spend over the full price of every package those leads signed — before studio costs, and before the collections book takes its share of what is never actually paid. A band sitting near 100% is not breaking even; it is a long way behind.

2. "On the calendar" is a SECOND independent join, not a funnel step under bookings. It matches the same lead against the Booked Client Data roster, which is a different CRM feed from the lead-bookings report and covers a different set of rows. The two do not nest: a band can carry more calendar rows than bookings, and the 45+ bands here do. Read it as a second opinion on how much real appointment volume a band produced, never as "of those bookings, this many showed". Neither column is a show rate — that lives on https://reports.iconicbyai.com/distance-age, where the roster is the denominator it belongs to.

3. The three axes are not independent. Age, ZIP and distance move together — the far bands really do skew older. Two tabs guard against reading one as the other: the Location tab age-standardises every ZIP prefix, and the Distance tab prints the age mix inside each band plus the full band × age rate grid. The cut scenarios carry no such guard: cutting 55+ and cutting past-150-miles overlap heavily, and their savings cannot be added together.

4. Distance is straight-line miles from a ZIP CENTROID. It is the distance to the middle of a postcode, not to a front door, and it is a crow-flies line, not a route. For anything where the road matters, read the drive-time table on the Distance tab instead — it uses the studio's own prefix sheet. 99.0% of leads are geocoded; the rest have no ZIP on the record or a ZIP the geocache does not carry, and they sit in their own band rather than being dropped or guessed at.