skip to content

For user research on a B2B analytics dashboard, how do you recruit the right participants and write a screener that does not give answers away?

level: middleimportance: should knowfreq 32%

answer

  1. criteria come from the question
  2. behaviours, not demographics
  3. buyer, admin and daily user differ
  4. hide the target answer among options
  5. beware the friendly-customer sample

basics

~20 s

Derive participant criteria from the research question — behaviours, role and context rather than demographics — then screen with multiple-choice questions that hide which answer qualifies. Recruit beyond friendly customers, and plan for no-shows and B2B incentive rules.

solid answer

~50 s

Recruiting starts from the **research question**: which **behaviours and contexts** must participants have — say, analysts who built or abandoned a report in the last month — and which segments must be covered, such as new and long-standing accounts, or churned ones. In B2B, the **buyer**, the **admin** and the **daily user** are different people, so name which you need. A **screener** is the short questionnaire that checks candidates against those criteria, written so it does not reveal the qualifying answer: multiple-choice questions with plausible decoy options and frequency ranges rather than yes/no, the study topic kept vague, exclusions for employees, competitors and recent participants, and one open question to check the person can articulate their experience. Recruit through several channels, not only account managers who tend to offer happy customers, and over-recruit for no-shows.

go deeper

for a junior

Recall what a screener is for and that participant criteria come from the research question, based on behaviour and role rather than demographics.

for a middle

Write screener questions that hide the qualifying answer — decoy options, frequency ranges, a vague topic, an open articulation check — and name the B2B roles to recruit.

for a senior

Diagnose recruiting bias in a finished study — friendly customers, only active users, professional participants — and plan channels, exclusions and over-recruiting to counter it.

for a principal

Build recruiting as an ongoing capability — a consented customer panel, fair incentive rules, participation caps — so studies are not held hostage to whoever is reachable.

## Why recruiting decides the result The best-run interview with the wrong person produces confident, useless findings. **Recruiting** — finding and selecting participants — is where many studies go wrong without anyone noticing, because the sessions still feel productive. ## Criteria come from the question For a study on why analysts abandon a B2B analytics dashboard's report builder, the criteria follow from the question: - **Behaviour**: started at least one report in the last month; include some who finished and some who abandoned. - **Role**: the analyst who builds reports, not the executive who reads them, unless the question covers both. - **Context**: account size, industry and how long the account has used the product, if those plausibly change behaviour. - **Exclusions**: your own employees, competitors, and people who took part in a study recently. Demographics such as age or location rarely matter in B2B research unless the question makes them matter; behaviours and context almost always do. ## Buyer, admin and daily user In B2B products, three different people are often lumped together as the customer: | Person | Relationship to the product | What they can tell you | |---|---|---| | **Buyer** | Chose and pays for it | Why it was bought, what success means to the business | | **Admin** | Configures accounts and access | Setup, permissions, governance pain | | **Daily user** | Uses it to do their job | Workflows, workarounds, friction | A study about report building needs daily users; interviewing buyers instead gives a confident picture of a workflow they never perform. ## Writing a screener that does not leak A **screener** is the short questionnaire candidates complete before being invited. Its danger is that people who want the incentive, or want to be helpful, will guess the qualifying answer. 1. **Avoid yes/no questions** about the target behaviour. "Do you build reports weekly?" signals the answer. 2. **Use multiple choice with decoys**: "Which of these tasks have you done in the last month?" with report building among several plausible tasks. 3. **Ask frequency as ranges**, not as a single qualifying value. 4. **Keep the study topic vague** in the invitation: "a study about how teams work with data", not "a study about our report builder's problems". 5. **Add one open question** — "Describe the last report you built" — to check the person has the experience and can talk about it. 6. **Screen out** employees, competitors and frequent research participants. ## Channels and their biases - **Account managers and customer success** know the customers but tend to offer satisfied, friendly accounts. - **In-product invitations** reach active users, not the churned or those who never adopted. - **Customer panels** are convenient but over-represent enthusiasts. - **Support tickets** find people with problems, but only those who complain. Mixing channels, and deliberately including churned and low-activity accounts, counters each bias. ## Numbers, incentives and no-shows - **Sample size**: a common practice for interviews is roughly five to eight participants per distinct segment, stopping when new sessions add little. It is a rule of thumb, not a standard. - **Over-recruit** by one or two per segment to cover no-shows. - **Incentives** in B2B may be restricted by the participant's employer; alternatives include charity donations, early access or a summary of the findings. - **Consent and data handling** are agreed before the session, including whether customer data may be shown. ## Checking recruits before the session - **Confirm high-stakes recruits** with a short call or message that asks the open question again. - **Track who took part** across studies, so the same customers are not asked repeatedly and research fatigue does not strain the relationship. - **Replace mismatches early**: a participant who turns out not to fit is better cancelled politely and paid than interviewed and quietly excluded, which wastes a slot and tempts the team to use the data anyway. Good recruiting is slow and unglamorous, and it is the difference between findings about your users and findings about whoever was easiest to reach.

  • Why include churned or low-activity accounts when studying report-builder abandonment?
    Because they are the people the problem affected most. Active, satisfied users have already found ways around the report builder's gaps, or never hit them. Churned and low-activity accounts are more likely to show where the product failed them. Leaving them out biases the study towards people for whom the product already works.
  • What is the risk of professional research participants, and how do you screen them out?
    People who take part in many studies learn what researchers want to hear and how to pass screeners, so their answers become rehearsed and less representative. Screen them out by asking when they last took part in a study, capping participation frequency in your own panel, and using an open-ended question that exposes vague or generic experience.

saying these in an interview costs you the question

  • A yes/no screener question about the target behaviour is the clearest filter.
  • Account managers' favourite customers are the most useful participants.
  • Demographics matter more than behaviour when choosing B2B participants.
  • The buyer can speak for how daily users do their work.
  • Telling candidates the exact study topic improves screener accuracy.