What a participant pool is and is not
The customer research programme playbook keeps recruitment standing because finding people is the slow part of research. Its pool contains only people who said yes to research contact, each with a note of what they accepted and when, and its health check asks three things: are the consent records current, are the same voices heard too often, and does the pool still look like the customers the decisions are about. The playbook gives the pool only a few lines. This tutorial is the build behind it: how people come to join, what they are told, which fields the pool keeps, how it stays apart from marketing, how invitations are spread and logged, and how the whole thing is reviewed.
A pool is a source of participants, not a sample and not a substitute for recruitment. Every study still describes whose experience answers its question, screens candidates against that description and invites a few of them. The customer research section gathers the methods those studies use, from one-to-one customer interviews to surveys and diary studies. The pool shortens the search; it does not decide who suits a particular question.
It is also not one of the standing groups that customers join for other reasons. A customer advisory board is a group that agrees to advise over time and argues about plans with the business and with each other; its own guide is clear that a board is no representative research sample. A beta or early-access programme gives chosen customers a working version of the product before launch, on terms written for that purpose. Each is its own commitment, made on its own terms, and the table shows why none of them can stand in for a pool.
| Group | What people agreed to | What the business uses it for | What it cannot do for research |
|---|---|---|---|
| Research participant pool | To be invited to research, and free to say no | A place each study screens and invites from | Stand for every customer, since it holds only people willing to be asked |
| Customer advisory board | To advise as a group over time, under a charter | Debating direction and work in progress | Stand for customers at large, since members are chosen and well informed |
| Beta programme | To try a working product before launch on stated terms | Learning whether the product works and what the launch needs | Supply neutral participants for other studies, since members carry one launch's terms and expectations |
| Marketing list | To receive promotional messages | Campaigns | Permit research invitations, since accepting marketing says nothing about research |
Before you start
Have the method for recruiting a single study in hand. The tutorial on recruiting research participants describes people by what they did and when, chooses a frame and writes down whom it misses, screens with neutral questions, invites on a research footing and keeps a recruitment log apart from the findings. A pool replaces only the search for names. Everything else in that tutorial still happens for every study drawn from it. That includes a recruitment logThe study's restricted record of who was approached, who qualified and took part, with screener answers, consent, sessions, incentives and a deletion date. for each study, kept separately from the pool's own records.
Find whoever owns data protection in the business and bring them in before any route opens. The tutorial on participant information and consent lists every use a study makes of what people say and gives each use its own consent statement, one of which is contact about future research. A pool turns that one use into a standing arrangement, so its basis, how long its records are kept and the words joiners see are that person's decisions, settled before anyone joins. This page describes general practice and what the site's guides say; it is educational and is not legal advice.
Find the contact rule the business already applies, if it has one, and list every team that approaches customers for research, feedback or references. A pool that ignores those teams protects nobody from being asked too often.
Gather what you already hold: consent records from past studies, including who agreed to be contacted again and the exact wording they saw, and the customer fields your records keep reliably. A pool that nobody maintains can turn, quietly, into a list of people whose agreement nobody can show, so agree how much time the owner can give it.
The steps
Decide what the pool is for, who owns it and who may draw from it
Write a short paragraph saying what the pool is for: recruiting participants for the business's own research, and nothing else. Name the studies it serves, such as interviews and usability sessions, and the ones it does not, such as a survey that needs a draw from the whole customer list. Then write the roles down as a RACI matrix, whose rule of a single accountable role per row stops shared ownership turning into no ownership. The rows a pool needs are the joining wording, the basis and retention, who may request a draw, who runs each draw, the review and the deletions. Whoever owns data protection is accountable for the wording, the basis and retention; the research owner for the rest. Requests for participants come through the owner alone: the single owner and single route that the customer references and advocacy guide uses so that its most helpful customers are not asked by everyone at once.
Choose the routes people join by, each on a research footing
List the ways customers can join and write down, for each, whom it reaches and whom it misses. Routes can include the consent statement at the end of a study asking whether the participant may be invited again, a research sign-up in the customer's account, and an invitation in the product or on a page customers reach after a particular event. Each route reaches some customers and not others: an account sign-up reaches people who log in, and an end-of-study statement reaches people who already took part. Keep every route about research only. The email guide that sets promotional against transactional messages shows that a message's purpose, rather than what it is called, decides which kind it is, and the recruiting tutorial applies that to research: an invitation that promotes a product counts as marketing, however its subject line reads, so no offer, discount or product news goes with an invitation to join. Where the data protection owner has said that marketing opt-outs are to be left out, check each new joiner against the Suppression ListA list of addresses or contacts excluded from every send, such as unsubscribes, hard bounces and spam complaints. first, and record the decision either way.
Write what joiners are told and record what each person agreed to
Joiners need to know what they are signing up to before they sign. Say who runs the pool and for what; that joining means being invited from time to time and never means having to take part; what the pool keeps about them and where it comes from; that joining brings no sales contact; how long their record is kept; and how to leave. Those lines rest on the MRS and ESOMAR codes as the site's research guides describe them: taking part is a voluntary and informed choice, people know who is asking and why, what they say stays confidential unless they agree to something else, and research must never serve as cover for selling. The Lawful BasisOne of the six grounds the GDPR accepts for processing personal data, such as consent, contract or legitimate interests, chosen for each purpose. is chosen for this purpose before any record is created and noted beside it; as the glossary entry summarises the GDPR, consent, where that is the basis, has to be given freely, for a specific purpose, on an informed and unambiguous footing, and be as simple to take back as it was to give. The Retention PeriodHow long a kind of personal data is kept for a stated purpose before deletion, set with whoever owns data protection and applied in every tool holding a copy. belongs to the data protection owner. Then record, for every member, the route, the date and the version of the wording they saw, so the business can show what each person accepted. A member whose agreement cannot be shown is not a member.
Hold only the profile fields a draw needs, taken from records where possible
A profile field earns its place only if a draw uses it to find the right members for a study. Plan, product, tenure, region and how recently someone bought are the kind of fields a draw sorts on, and where the business's own records hold them, a draw can read them from there rather than from a joining form, whose stated answers date. Where the pool can look a field up from the customer record by an ID rather than copying it, do that. Ask joiners only for what the records cannot supply, such as whether they can join a video call. Every field is Personal DataAny information relating to an identified or identifiable living person, such as a name, email address, location data or an online identifier., so collecting less is the GDPR's data minimisation in practice. A profile is never a substitute for the ScreenerThe short questionnaire that confirms each research candidate fits a study, asking about recent behaviour without revealing which answers qualify.: profiles go stale, and a member who fitted when they joined may not fit a study now. Each study still confirms fit with its own neutral questions before anyone is booked.
Hold the pool as its own list, separate from marketing lists
Keep the pool in its own store, reachable only by the research owner and the people the RACI rows name, and do not sync it to any marketing audience, segment or automation. Joining the pool adds nobody to marketing, and an invitation to a study goes out as research, from a sender and in a format that make that plain. Separate lists also let a request to leave the pool stop research invitations without changing the person's marketing choices. Before each draw, apply the contact rules and the research footing as the customer-list sampling tutorial sets them out for its frame: people contacted for research within the rule's period are passed over, whichever team contacted them, and the opt-out decision the data protection owner made is applied every time rather than once at joining.
Rotate invitations across members so the same people are not drawn each time
Members who reply quickly and say yes to everything can become the people every study hears from unless the draw stops it. For each study, filter the pool to members who could fit the description, remove anyone protected by the contact rule, and select at random among the rest rather than taking the first names, the newest joiners or the most willing. Members invited recently, whether or not they took part, wait until others have had a turn. The voice of the customer guide applies the same thinking to feedback surveys: a customer asked recently is left alone for a period, whatever team wants to ask. Where the customers are businesses, a few key accounts can be asked for research, references and advisory input by different teams at once, and the guide to researching business buyers tracks, account by account, who has been asked, for what and when, so that a study from one team does not arrive on top of a request from another. Write down how each draw was made, so anyone can check it was spread fairly.
After each draw, note which customers the pool misses
Every draw shows the pool's shape. Compare the members a study needed with the members the pool could offer, and write a short note for the owner: which kinds of customer the study wanted and could not find, and which kinds filled the places instead. The pool holds people who were around and willing when a route reached them, which is the filter the survivorship bias framework warns about: customers who lapsed, paused or never engaged are missing, and they may be the people a question is about. The members who joined may also differ from the customers who saw a route and did not, a close relative of Non-Response BiasThe gap between the people a survey or study invited and those who answered; engaged, satisfied customers tend to reply more, so results lean their way. that only shows when the pool is compared with the records. When a study needs people the pool lacks, recruit them for that study from another frame using the recruiting tutorial's method, and say in the findings where the participants came from. When the same gap appears draw after draw, add a joining route that reaches the missing group, still on a research footing.
Keep a participation record for each member
Keep one record per member of what the pool has asked of them: each invitation with its date and study, whether they took part, did not reply, declined or did not turn up, any incentive paid, and any withdrawal. The record is what makes rotation and the contact rule work, and it is how the owner answers a member who asks what has been done with their details. Keep it with the pool, not with the findings. Each study still keeps its own recruitment log and gives participants codes, and the research session readiness checklist confirms that log and the consent documents before a moderated session starts. The findings themselves go to the store that building a research repository sets up, which holds findings and de-identified excerpts and leaves contact details, screener answers and recruitment logs out. Nobody reading a finding should be able to work back to a pool record, and nobody reading a pool record should see what the member said.
Review consent, leavers and the pool's make-up on a fixed date
Set the review date when the pool opens, and put it in a calendar that will outlast the people who set it up. At each review, check that everyone who asked to leave has gone and that the request stopped invitations at once, since leaving should be as easy as joining. Apply what the joining wording said about members whose customer relationship has ended; whether former customers stay is decided at the start and written into what joiners are told, because keeping them helps studies of why people leave only if they agreed to it. Remove records that have reached the deletion rule the data protection owner set. Check that every remaining member has a recorded route, date and wording version. Then lay the pool's make-up beside the customer records on the fields the draws use, read the gap notes since the last review, and decide which routes to add or close. Record what changed and set the next date. The playbook's health check on consent, repeat voices and fit is the review's agenda.
What a pool record holds
The table lists what one member's record holds, where each field comes from and what it lets the owner do. It holds no answers from any study.
| Field | Where it comes from | What it lets the owner do |
|---|---|---|
| Member code | Assigned at joining | Refer to the member without a name in draws and gap notes |
| Contact detail for research invitations | Given at joining or held on the customer record | Invite the member, and nothing else |
| Route, date and wording version | Recorded at joining | Show what the member accepted, when and how |
| Uses accepted | The member's choices at joining | Invite only to the kinds of study the member agreed to |
| Profile fields | Looked up from customer records where possible | Filter for studies without asking again |
| Invitations and studies | Added at every draw | Apply rotation and the contact rule, and answer the member's questions |
| Withdrawal and its date | Recorded the moment the member leaves | Stop invitations at once and show that it happened |
| Deletion rule | Set by whoever owns data protection | Remove the record when its time is up |
The invitation and study fields also speed up the review: a member with a long run of invitations and no studies, or study after study in a short time, stands out at once.
A flower delivery business builds its research pool (illustrative)
Vlodmerra Flowers builds a research pool from its subscribers
Vlodmerra Flowers, an invented flower delivery business, sends regular deliveries to subscribers on weekly, fortnightly and monthly plans. Its last 3 studies each recruited from scratch. The research owner sets up a pool, with the privacy lead accountable for the joining wording, the basis and the deletion rule, and opens two routes: a statement at the end of each study asking participants whether they may be invited to future research, and a research sign-up in account settings that says the pool exists for research only and promotes nothing.
Of 46 past participants, 29 agreed to future contact in wording the team can still produce; the other 17 are not added. In the first two months, 212 subscribers join from account settings. Each record holds the route, the date and the wording version, and plan, tenure and delivery status are looked up from the subscription records by customer ID instead of being asked. The pool sits in a restricted store with no link to the email platform, and joining puts nobody on a marketing list.
The first study, interviews about choosing delivery days, needs 8 weekly subscribers. The owner filters the pool to 97 weekly members, passes over 5 surveyed by the service team within the business's contact rule, and draws 16 at random; the screener confirms fit before anyone is booked. The next study draws from members not invited the first time.
The gap note after the first draw compares the pool with the subscription records: subscribers who have paused deliveries make up 19% of subscribers but 4% of the pool, because the account-settings route reaches people who log in, and paused subscribers had little reason to. For a study about pausing, the team recruits from the paused list with its usual recruiting method, and adds a third route, a research-only invitation on the page that confirms a pause.
Each member's participation record lists invitations, studies, no-shows and incentives paid. At the review date set when the pool opened, the owner removes 23 members who closed their accounts, as the joining wording said it would, and 6 who withdrew, and confirms that each withdrawal had stopped invitations the day it arrived. Paused subscribers now make up 11% of the pool.
Common mistakes
- Filling the pool from the marketing list. People who accepted promotional messages never agreed to be invited to research, and their records cannot show that they did.
- A joining route with an offer attached. An invitation that promotes something is marketing, and it fills the pool with people who came for the offer.
- No record of what each member accepted. Without the route, the date and the wording version, the business cannot show why a member is in the pool.
- The pool synced to marketing tools. Once the lists are joined, research members start receiving campaigns and the purposes can no longer be told apart.
- Drawing the quickest responders. Taking whoever replies first turns a pool into a small panel of enthusiasts who answer every study.
- Treating the pool as the customer base. It holds the willing and the reachable; each draw should end with a note of whom it could not supply.
- Advisory boards and beta groups as the default source. Those groups have their own terms and their own stake, and drawing on them for every study narrows what research hears.
- No review date. Leavers, expired records and a drifting make-up pile up until nobody trusts the list.
