A five-step workflow for estimating who will qualify for a B2B study — and how long it will take to reach them — before field begins. Built on treating incidence as a modelled range to be validated, not a published benchmark to be looked up, and grounded in the AAPOR, ESOMAR, and ISO 20252 standards buyers increasingly expect.
CatalystMR Research Team. (2026). B2B Sample Incidence — A Methodology for Modelling Feasibility by Title and Industry. CatalystMR Methodology Papers. https://www.catalystmr.com/insights/methodology-papers/b2b-sample-incidence/
@techreport{catalystmr_b2b_sample_incidence,
author={{CatalystMR Research Team}},
title={B2B Sample Incidence — A Methodology for Modelling Feasibility by Title and Industry},
institution={CatalystMR}, year={2026}, type={Methodology Paper},
url={https://www.catalystmr.com/insights/methodology-papers/b2b-sample-incidence/}
}TY - RPRT AU - CatalystMR Research Team TI - B2B Sample Incidence — A Methodology for Modelling Feasibility by Title and Industry PB - CatalystMR PY - 2026 UR - https://www.catalystmr.com/insights/methodology-papers/b2b-sample-incidence/ ER -
Incidence rate — the share of an accessible target audience that qualifies for a study — is the single variable that most often decides whether a B2B project comes in on time, on budget, and on specification. Yet it is routinely treated as a fixed number to be looked up, when in reality it shifts with seniority, function, company size, industry, geography, and decision authority, and again with every change to the screener.
Rather than offer benchmark tables, this paper gives a repeatable five-step workflow for modelling B2B incidence honestly — define the criteria, model the range, translate it into a fielding plan, validate with soft-launch, and revise — and then maps the levers that move a modelled incidence up or down. It deliberately publishes no incidence figures, because the right number is the one modelled from your audience, in your market, against your finalised screener.
In survey research, incidence rate is the percentage of people in the accessible target population who meet a study's qualification criteria — a close cousin of the qualification and eligibility rates that the AAPOR Standard Definitions codify for the field.2 A broad B2B study may qualify managers across many industries; a narrow enterprise-software study may require only budget owners at companies above a specific revenue threshold. The difference between those two is not a detail — it is the difference between a routine fortnight in field and a project that quietly fails to launch.
Incidence is not just one input among many; it sets the ceiling on what is feasible. It drives screening effort, the incentives required, field time, the source mix, the total sample you must touch, and ultimately cost. A study's design can be flawless, but if its real incidence is a fraction of what was assumed, every other plan built on top of it gives way.
The most common — and most expensive — mistake is to treat incidence as a fixed figure to be retrieved from a table and dropped into a plan. Published "benchmarks" feel authoritative, but a benchmark drawn from a different audience, market, or screener tells you very little about your study. Incidence is a property of a specific question asked of a specific population at a specific moment; it is modelled, not looked up. The rest of this paper sets out how to model it — as a five-step workflow.
If a feasibility number arrives without the assumptions behind it — the audience, the market, the screener it was modelled against — it is not yet a feasibility estimate. It is a guess wearing the costume of one.
Because incidence is a property of a specific study rather than a fixed fact, it has to be built — through a repeatable sequence that ends in a measurement, not an assumption. The five steps below are the spine of this paper; Sections 03–05 take them in turn, and Section 06 maps the levers that move the number up or down.
Capture title, function, firmographics, and behaviour as explicit, separate variables.
Estimate a banded incidence — upper and lower — tied to the finalised screener.
Convert the range into field time, source mix, sample volume, and cost.
Release a small fraction to measure the real qualifying rate before committing.
Re-model against measured data; trigger backups or a method change as needed.
A benchmark answers "what is incidence for this kind of audience?" — a question your study has already outgrown the moment it is specified. The workflow answers the only question that matters: "what is incidence for this audience, against this screener, in this market — and how confident are we?" Each step narrows that uncertainty deliberately, and the final two steps replace estimate with evidence.
The workflow is drawn left to right, but Steps 4 and 5 feed back into the plan. Soft-launch almost always teaches you something the model could not — a qualifying rate slightly off, a cell filling slowly — and the discipline is to revise the plan against that evidence before full commitment, not to push on and hope the average recovers.
The first two steps decide the quality of everything after them. Step 1 makes the target explicit; Step 2 turns it into a banded estimate. Incidence moves on three axes — title, firmographics, and behaviour — and a target's true incidence is the product of all three narrowing at once, not any single one.
Seniority (Manager, Director, VP, C-Suite) and the functional lens — finance, IT, security, marketing, HR, operations.
Employee count, revenue band, geography, and industry vertical. The same title means different things across these.
Software usage, purchasing stage, installed base, recent purchase, and purchase involvement or authority.
Each criterion you add narrows the qualifying population, and they compound. A "Director of IT" audience may be readily feasible in online panel; CISOs at hospitals using a specific cybersecurity platform stack four constraints — seniority, function, industry, and a behavioural installed-base requirement — and may require telephone (CATI) outreach and partner sourcing to reach at all. This is why a single headline figure for "B2B decision-makers" is meaningless for planning: the moment you specify the study, you are no longer sampling that population.
The same audience can yield very different incidence depending on how the qualifying question is worded. "Do you influence IT purchasing?" and "Are you the final decision-maker for IT security tooling?" describe overlapping but very different populations. Until the screener is locked, any incidence figure is provisional by definition — which is why Step 2 produces a range with its assumptions attached, re-modelled once the screener is final. The ESOMAR framework explicitly asks providers what they do to put upper and lower boundaries around feasibility.1
Illustrative only — relative widths show how added criteria compress the qualifying base. CatalystMR publishes no fixed incidence percentages; the figure for any study is modelled from that study's own audience, market, and finalised screener.
A modelled incidence is only useful when it is translated into a fielding plan. The banded estimate from Step 2 is used to project field time, source mix, sample volume, and cost before launch — and to decide, for low-incidence audiences, whether online panel alone can carry the study or whether it needs reinforcement.1
For low-incidence B2B audiences, online panel alone may not reach the depth a study needs. This is the point at which a feasibility model should recommend telephone (CATI) interviewing, screen-sharing CATI for visual tasks, mixed-mode recruitment, or partner sourcing — not as upsells, but as the practical means of filling cells that online cannot.
Ask directly: "At this incidence, can online panel alone deliver — and if not, what's the plan?" A credible answer names the method change before launch, not after the field stalls.
A large B2B panel can still be thin for a narrow, low-incidence target. What matters for your study is not a total count of business professionals but the verified qualifying depth in your specific audience. Headline panel numbers tell you almost nothing about whether a narrow, behaviourally-defined cell can actually be filled — which is why incidence must be modelled against your criteria, not inferred from a provider's size.
A modelled incidence is a hypothesis. Soft-launch is the experiment that tests it, and revision is what you do with the result. Releasing a small fraction of sample first converts an assumption into a measurement — so that if real incidence tracks the model, the full launch proceeds with confidence, and if it does not, you revise before the budget and timeline are committed rather than after the field is contaminated by the scramble to hit a number.
| Consideration | Without soft-launch | With soft-launch |
|---|---|---|
| Incidence assumption | Discovered wrong in full field | Measured on a small release first |
| Quotas & targeting | Adjusted mid-field under pressure | Tuned before full commitment |
| Backup audiences | Improvised when the cell stalls | Named and ready in advance |
| Budget exposure | Committed before evidence | Released against measured incidence |
The time to decide how you will respond to a low qualifying rate is before launch, not during it. A sound plan names its fallbacks in advance: relax the seniority floor, broaden the industry or behavioural criteria, add a telephone channel, or bring in partner sourcing — each chosen so the research objective survives the adjustment. Backups improvised mid-field, under deadline pressure, are how a study quietly drifts off its specification.
Soft-launch is not a delay; it is the cheapest insurance in the project. A small measured release almost always costs less than a full field built on an incidence assumption that turns out to be wrong.
Incidence is not a fact you discover; it is the net result of choices you make. Each lever below is a decision the study controls, and moving it tightens or loosens the qualifying base in a predictable direction. The settings are deliberately directional, not numeric — the magnitude is what Step 2 models and Step 4 measures for your own audience and screener.
B2B incidence rewards planning at the front of the process and punishes assumptions at the back. Run the five steps every time: define the criteria across title, firmographics, and behaviour; model a banded range tied to the finalised screener; translate it into field time, source mix, and cost; validate it with soft-launch; and revise against the evidence before committing. Each step is independently sensible, and together they convert a hard-to-plan audience into a fielding plan you can stand behind. The AAPOR, ESOMAR, and ISO 20252 standards exist precisely so buyers can ask for this rigour in consistent terms — and recognise it when they see it.1,2,3
CatalystMR is a global market-research panel and fieldwork partner specialising in hard-to-reach B2B, executive, and niche audiences. We model feasibility before field and pair verified online sample with live telephone (CATI) capability for the low-incidence targets that online alone cannot reach — under one screener, one QC standard, and one point of contact.
We publish our own responses to ESOMAR's 37 Questions and treat incidence as a modelled range to be validated by soft-launch, not a benchmark to be promised.
Compliance posture: our methodology is aligned to the ESOMAR Code and Guidelines and the ISO 20252 framework, and we are certified under the EU–U.S., UK, and Swiss Data Privacy Frameworks, with personal data siloed from response data.