Funnel velocity model

What a faster, more decisive sales cycle is worth: today's real HubSpot numbers, next to a stronger-brand scenario you set yourself.

Working model · not a forecast
01 · the real funnel isn't one lane, it's two

Most deals never go through a formal evaluation. The ones that do are worth nearly double.

Pulled fresh from HubSpot (Sales pipeline, New Business, the 621 deals closed won in the trailing 12 months to 15 Sep 2026): three in five close in days, with no evaluation step at all. The other two in five sit in a genuine evaluation stage, take about five times as long, and are worth 70% more on average.

Fast lane

370 deals · 59.6%
Median 4.2 days lead to close · average deal £5,444

Evaluated lane

251 deals · 40.4%
Median 20.8 days lead to close · average deal £9,336

This is the lever a stronger brand actually pulls. Not a uniform speed-up across every deal, but a mix shift: more buyers arrive already trusting Joblogic enough to skip the formal evaluation, and the ones who still need it move through it faster because there's less convincing left to do.

02 · how the numbers are built

Two lanes, weighted together

Leads/mo×Win rate=Customers/mo → split byEvaluated mix %→ each lane × its ownACV=New ARR/mo

Days-to-close works the same way: each lane has its own cycle time, weighted by the same mix, to give one blended "days-to-close" figure per scenario.

Worked example, today's real numbers: 228 leads/mo × 20.2% win rate = 46.1 customers/mo → 59.6% fast lane (27.5 × £5,444) + 40.4% evaluated lane (18.6 × £9,336) = £323,372/mo new ARR, blended 10.9 days to close.
03 · where each number comes from

Every default is real, not assumed

InputToday's valueSource and how solid it is
Leads/month2282,735 New Business deals created in the Sales pipeline (HubSpot portal 27061482), trailing 12 months to 15 Sep 2026, ÷ 12. A deal-level count, not a contact-level one, so it won't match agentic-business-case's 257/mo lead figure exactly, but it's the right denominator for this pipeline.
Win rate20.2%553 of those 2,735 are Closed Won. Close to the 20.7% found in the June/September funnel baseline, cross-checking cleanly.
Evaluated mix40.4%Of the 621 deals closed won by close date in the same 12 months, 251 have a timestamp for entering "Business Case/Evaluation"; 370 don't. This is the thinnest input conceptually, not statistically: it's a real stage flag, but a rep's habit of logging the stage move, not a hard rule the CRM enforces.
Fast lane days-to-close4.2Median, the 370 fast-lane deals, create-date to close-date. Median not mean, because a handful of multi-month outliers would otherwise drag the average up.
Evaluated lane days-to-close20.8Median, the 251 evaluated-lane deals, create-date to close-date. Splits further into a median 2.7 days to reach evaluation and 13.0 days inside it, if you want the finer detail.
Fast lane ACV£5,444Mean deal value (amount in home currency, GBP), the 370 fast-lane deals.
Evaluated lane ACV£9,336Mean deal value, the 251 evaluated-lane deals. Weighting both lanes by mix reproduces the pooled average deal value (£7,017) almost exactly, which is a good sign the split is measuring something real rather than noise.
04 · try your own numbers

Today, next to a stronger-brand scenario you define

This isn't a sensitivity band. It's two scenarios, side by side, and you set both. The left column starts on today's real figures. The right column starts on an illustrative stronger-brand case, a bigger win rate, more deals skipping evaluation, evaluation itself moving faster, all overwritable. Nothing here is a prediction; it's a place to put your own hypothesis and see what it's worth.

Held constant on purpose: this isolates the conversion and velocity effect from traffic volume, which the branded search and sitewide CVR models already cover. Change it here if you want to stack a volume assumption in too, but check you're not double-counting against those two.

Today

Real baseline
Customers/mo
-
Blended ACV
-
New ARR/mo
-
Days to close
-

Stronger brand

Illustrative, overwrite me
Customers/mo
-
Blended ACV
-
New ARR/mo
-
Days to close
-
05 · what that's worth

The delta between the two columns

Extra customers/mo
-
Extra new ARR/mo
-
Cumulative, 12mo
-
Faster to close
-

Cumulative new ARR, 12 months, held flat (gross bookings, no churn)

Today Stronger brand
  1. This model holds monthly lead volume constant across both columns on purpose, so what you're seeing is the effect of conversion and speed alone. If you also want to model more leads, use the branded search or sitewide CVR models and add the two deltas carefully; the leads pool overlaps, so simply adding all three ARR figures together will overstate the case.
  2. The "stronger brand" column starts on illustrative numbers, not a prediction. Nothing about the win-rate lift, the mix shift, or the faster evaluation is measured yet; it's a placeholder for Dan's own hypothesis or the brand strategist's evidence once it exists.
  3. The evaluated-mix split is a real HubSpot stage flag, but depends on reps consistently moving deals into "Business Case/Evaluation" rather than skipping the click. Treat 40.4% as directionally right, not exact to the decimal.
  4. Days-to-close uses medians per lane, which is right for typical-deal behaviour, but the mean cycle time across all 621 deals is much higher (32 days) because a small number of deals take months. Don't quote the median as "the sales cycle" without this caveat.
  5. Gross bookings only: no churn, no contraction, no compounding traffic growth. Cumulative ARR is new ARR/month held flat and multiplied by the month count, same convention as the branded search and sitewide CVR models.
  6. Before this goes anywhere near Emma, Katarina, Jim or Phil, the "stronger brand" inputs need to be either evidenced or clearly relabelled as a working hypothesis, not left as this build's placeholder numbers.
Built 2026-09-16 by Dan's marketing team · inputs are live, drag or type to re-run · questions on the method go to Dan