Funnel leak calculator for healthcare and wellness teams
Your traffic, your conversion rate, your order value. It works out what the drop is costing you each month, and which step I’d look at first. No email, no form — the numbers never leave your browser.
Your funnel
Your own figures, multiplied out
There are analytics and you don’t trust them, which in practice is the same as not having them. Fixing that is cheaper than arguing about what a rebuild did.
At this volume most single-step changes can be read honestly within four to six weeks.
Not a forecast — a range to think inside. The low end is deliberately cautious. The high end sits near the best lift I’ve measured on a rebuild, against a baseline agreed before anything shipped. Which end you land nearer to is decided by your funnel, not by this page.
Four ways in
The same funnel, audited, patched, rebuilt or kept
What each one costs, what you get, and the lift it would have to produce to cover its own fee in the first month — worked out on the numbers you just entered.
01
Matches your answers
Funnel audit
- Fee
- From $1,800
- Timeline
- Two weeks, fixed fee
- Lift to cover it in month one
- +7.2%
- What you get
- The drops ranked by what each one costs, and a straight answer on whether a rebuild is worth it.
- How it is read
- Nothing ships, so nothing moves. It sets the baseline everything after it is read against.
- Where it is the wrong call
- If you already trust your numbers and know which step is failing, it is two weeks you did not need.
02
Intake and eligibility
- Fee
- From $3,500
- Timeline
- Three to five weeks
- Lift to cover it in month one
- +14.0%
- What you get
- One expensive step rebuilt — intake and eligibility, or pricing and checkout.
- How it is read
- Completion of that step, against the baseline agreed before it shipped.
- Where it is the wrong call
- If the loss is spread evenly across the whole flow, one step moves less than it looks like it should.
04
Website or product design
- Fee
- From $8,000
- Timeline
- Scoped per project — usually six to ten weeks
- Lift to cover it in month one
- +32.0%
- What you get
- The whole site or product flow, designed around one decision, with the measurement in place before launch.
- How it is read
- Site conversion against the pre-launch baseline, read over weeks rather than days.
- Where it is the wrong call
- If a single step is doing all the damage, this is an expensive way to fix it.
05
Ongoing optimization
- Fee
- From $3,500/mo
- Timeline
- Monthly, flat fee
- Lift to cover it in month one
- +14.0%
- What you get
- A standing slot in the queue: one request at a time, most back in 48–72 hours.
- How it is read
- Month over month, with a written record of what moved and what did not.
- Where it is the wrong call
- If nothing is instrumented yet there is nothing to optimize against, and this becomes redecoration.
Fees are the published starting fees on each service page. The lift column is that fee against your current monthly revenue, as a relative lift — your figures, nothing modelled. Recurring work is read monthly.
How this is worked out
Where every number comes from
No figure here is a guess, and none of them are mine except the fees.
- Customers and revenue
- Your visitors times your conversion rate, times your order value. Nothing else goes into it.
- One point of conversion
- One percentage point of your traffic at your order value — the unit most funnel changes actually move in.
- The lift range
- 5% is deliberately cautious. 20% sits near the best relative lift I have measured on a rebuild, against a baseline agreed before anything shipped. Neither is a claim about your funnel.
- Fees
- The published starting fee on each service page, unchanged. What your funnel needs is scoped after a call, in writing, before anything starts.
- Lift to cover it
- The fee against your current monthly revenue, expressed as a relative lift. Recurring work is read against one month of fee.
- What it leaves out
- Repeat purchase, lifetime value, seasonality, traffic mix and statistical significance. First order value only, so the figure is a floor rather than a best case.
- As of
- Approximate, in USD, October 2026. Nothing you type is sent anywhere or stored.
Everything above the lift line is your own figure multiplied out — no benchmarks, no industry averages, nothing modelled. The lift range is a question, not a projection: the low end is cautious, the high end is the one result on this site that was measured against a baseline agreed before anything shipped. Fees are the published starting fees for each service. Nothing you type here is sent anywhere.
Questions about the number
The arithmetic, the assumptions, and what this deliberately doesn’t tell you.
No. The fee shown is the published starting fee for that service, which is the same number on its own page. What your funnel actually needs is scoped after a call, in writing, before anything starts.
The low end is a cautious 5%, well under what I’d hope for from a rebuild. The high end sits near the best relative lift I’ve measured, held for weeks against a baseline agreed before anything shipped. Neither is a promise about your funnel — I haven’t seen it, and any number that claims otherwise is marketing.
Then every figure here is wrong by the same proportion, which is the point. If you can’t put a number in box 03 with a straight face, that is the finding — and it’s why the tool sends you to the audit when you say you don’t trust it.
No, deliberately. It uses first order value only, so the figure is the floor rather than the best case. If your repeat rate is strong the real number is higher, and that’s a conversation worth having with your own cohort data in front of us.
No. It runs entirely in your browser — there is no form, no submit, and no analytics event attached to these inputs. If you want me to see them, the email button fills a draft in your own mail client and you decide whether to send it.
Then the honest answer is that a split test won’t finish in a useful time, and I’ll say so. At low volume the work is session review, drop-off analysis and bigger single changes judged over a longer window — not a test that never reaches significance.
Second opinion
Think that number is wrong?
Send me what you’d have expected instead. I’d rather be corrected now than halfway through a rebuild.
Book a callOr email me