Applying sector benchmarks to diagnose funnel leaks

The slide says your evaluation-to-purchase rate is 10% while the sector runs near 30%, and somebody in the room wants to reorganise the field team by Friday. Before any money moves, settle one question: does the peer number count the same events yours counts? Most funnelfunnelThe customer journey from awareness to purchase, typically Awareness, Interest, Consideration, Decision, Action, with prospects narrowing at each stage.View full definition → gaps presented to medtech and diagnostics boards shrink or vanish once you reconstruct the other party's denominator. The ones that survive that test are worth real money, and that is where the fix budget goes.

Three judgements sit with you and nobody else: where to get reference numbers you can defend, how to tell a leak from a measurement artefact, and how to price the repair against the alternatives.

Where defensible reference data comes from

Rank sources by what they can be held to, not by how quotable they are.

• Audited peer filings. Listed diagnostics and device companies disclose segment revenue, sales and marketing spend, installed base or test volumes. Slow, public, and hard to fudge. Two or three close comparables beat any survey.
• Syndicated data. IQVIA sells prescription, claims and national sales audits. It is a vendor of the data under discussion, so the methodology note matters more than the headline. Its real value is denominators you cannot observe yourself: treated patient counts, procedure volumes, your share inside a defined territory.
• Consulting benchmark panels. ZS Associates does commercial benchmarking and field force sizing for life sciences, and sells the follow-on work, which is a reason to inspect the sample rather than discard the number. A panel of 14 companies, nine of them large pharma, tells you almost nothing about a $40M diagnostics business.
• Software baselines. Bessemer's cloud benchmarks are honest about what they measure, which is SaaS. Borrowing a 12 month payback norm from that world and applying it to a capital-plus-consumable device business imports the wrong cost structure entirely.

One filter, applied before you read the number: population, denominator, period. If the source will not give you all three, it is an anecdote with a decimal point.

Leak or artefact?

Run the recompute test before you accept any gap.

Restate YOUR metric under each plausible reading of the peer's definition.
If the resulting spread covers the reported gap, the gap is not evidence.

The artefacts that produce most false leaks in this sector:

• Unit of customer. One nine-hospital IDN counted as a single account, or as nine? That choice halves or multiplies conversion rateconversion rateThe percentage of visitors or prospects who complete a desired action (purchase, sign-up, contact form), calculated as conversions divided by total opportunities.View full definition → and per-account acquisition costacquisition costCustomer Acquisition Cost (CAC) is the total sales and marketing spend divided by the number of new customers gained in a period. It measures how efficiently you grow.View full definition → at the same time, in opposite directions.
• Stage entry. A peer may open an evaluation only at signature of a study protocol; you may open it when a lab director says "send us samples". Same word, different point on the funnel the mapping lesson lays out.
• Cost scope. Whether committee-facing clinical education, freight and placed instruments sit inside the acquisition cost the way the acquisition cost lesson defines it. A 40% swing here is ordinary.
• Capital lumpiness. An account that bought a $400,000 analyser in year one and $160,000 of consumables in year two prints 40% retention while behaving perfectly. Split capital, disposables and service the way the lifetime valuelifetime valueLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.View full definition → lesson separates them before comparing retention with anyone.
• Small n. Six contracts from 60 evaluations gives a 95% interval of roughly 5% to 20%. Even the optimistic end sits under a 25% to 40% peer band, so that gap is real. Two from 12 gives an interval closer to 5% to 45% and decides nothing. Do not restructure a commercial team on twelve data points.

Exact Sciences and the stage no benchmark covers

Cologuard, approved by the FDA in 2014 with a CMS national coverage decision the same year, moves through a funnel that does not end at the order. A physician orders, a kit ships, the patient collects a sample at home and posts it back, the lab runs it, and only then is there revenue. Exact Sciences reports test completion because a kit that never returns is a shipped cost and nothing else, and the returned share has run in the region of two thirds.

No syndicated source benchmarks that stage. It exists because of the product's format, so the only reference series available is your own, cohort by cohort. That has two consequences a leader has to hold at once. Demand generationDemand generationMarketing activities designed to attract and capture contact information from prospects interested in your offer, creating a pipeline of potential customers.View full definition → that lifts orders also multiplies the downstream loss: roughly 300 of every 1,000 extra orders fall into the gap. And the arithmetic on the marginal dollar inverts. On 100,000 orders, moving completion by six points yields 6,000 additional resulted tests with no new demand spend at all. Buying those same 6,000 tests through advertising means funding about 9,000 orders at whatever your cost per order is. Reminder calls, navigator outreach and kit design almost always win that comparison, and almost never win the argument in the room, because the completion stage has no peer number to shame it with.

Deciding where the money goes

Value of closing a leak = revenue sitting behind the stage
                          x points of movement you can defend
                          x probability the fix works
                          - cost of the fix, and months until it bites

Check upstream definitions first, then spend downstream. A leak late in the funnel is more expensive per unit because the acquisition cost is already sunk, and it is usually cheaper to fix because the customer is already talking to you.

A worked arbitration

A diagnostics business runs 400 lab evaluations a year and shows three gaps: conversion at 10%, payback at 34 months, retention apparently at 88%.

Recompute first. Retention at 88% turns out to be a capital year rolling off; on the consumables stream alone the same accounts grew. That gap is an artefact and gets zero budget. Peers open evaluations at signed protocol, which lifts the honest peer figure down to about 22%. Real gap: 12 points on 400 evaluations, roughly 48 contracts at a $120,000 first year value, so $5.8M of trapped revenue. The proposed fix, hard qualification criteria plus redesigning the evaluation so it measures turnaround time rather than analytical concordance, costs about $600,000 and takes two quarters.

The alternative on the table was cutting field coverage to pull payback under 30 months, saving $1.5M of spend but removing the reps who run the evaluations. Funding the conversion fix and holding the field flat is the defensible call, and it is defensible precisely because two of the three gaps were shown to be definitional before anyone voted.

Failure modes worth naming out loud

Benchmark shopping. Sourcing until you find the range that supports the decision you already made. The countermeasure is procedural: name the source and the definition before you see the comparison, and write both into the board pack.

Goodhart at field level. Publish evaluation-to-purchase as a rep target and the ratio improves within two quarters while total contracts fall, because the safe evaluations are the ones that get started. Pair every ratio you publish with its absolute numerator and denominator on the same line.

Undated numbers. The payback tolerance of the 2021 funding market is not the 2025 to 2026 posture, where capital efficiency is priced. A benchmark without a year attached is a benchmark you cannot argue with, which is different from one you can trust.

Regulatory stalls read as marketing failure. An account frozen pending 510(kkThe average number of new users each existing user generates through referrals. Above 1.0, growth compounds on itself and becomes exponential.View full definition →) clearance or an IVDR-related transition is not a conversion problem, and re-engineering the pitch will not move it. Hold those deals in a separate pool and report the funnel without them, or you will keep fixing the wrong stage.

Key takeaways

• Reconstruct the peer denominator before you accept a gap. If restating your own metric under plausible alternative definitions produces a spread wider than the gap, you have no finding.
• Rank sources by accountability: audited filings, then syndicated data such as IQVIA, then consulting panels such as ZS Associates, and say plainly when the source sells the thing being measured.
• Your own trailing series is often the strongest evidence you have, and for product-specific stages such as Cologuard's kit return it is the only evidence that exists.
• Price the fix, do not just rank the gaps: revenue behind the stage, defensible movement, probability of success, cost and time to effect.
• The measurable leak and the expensive leak are rarely the same one. Downstream stages carry sunk acquisition cost, which makes a point of recovery there worth more than a point at the top.