Survey Screener Questions on LinkedIn: Qualify Respondents Before You Send the Link

Every paid survey with strict criteria has the same leak: people who don't fit open the link, fail the screener (or worse, guess their way through it), and you pay for answers you can't use. On LinkedIn there's a better order. Ask the screening questions in the conversation, and only send the survey link to people whose answers fit.

Short answer: invite people who look right on paper, ask whether they'd be open to a couple of quick questions to check the fit, then ask your screener one question at a time in chat. If they qualify, they get the survey link with the length and reward. If they don't, they get a warm thank-you and never see the link. Updately's Screen first, then send the survey template does exactly this.

Key Takeaways

Why screen in the chat, not just the survey

Most survey tools can screen people out on the first page, and you should keep doing that. Screening in the conversation as well adds things a form can't:

• The link stays private. It only reaches people who qualified, so it isn't shared around and your incentive budget isn't spent on misfits.
• Fewer awkward "sorry, you don't qualify" pages. Someone who gave you their time on LinkedIn deserves a human reply, not a terminate screen.
• Ambiguity gets resolved. "Sort of, I share that budget with IT" is a real answer. A form forces a box; a conversation can ask one follow-up.
• Quotas are easier to manage. When a cell fills, you change one line in Knowledge and the agent stops sending people from that group to the survey.

LinkedIn profiles get you most of the way, since title, company and industry are visible. Screening checks the things profiles don't show: who owns the budget, which tools they use, how recently they bought something.

Writing screener questions for a chat

Good screener questions are the same in a chat as in a form, with a few chat-specific rules:

1. Don't reveal the qualifying answer. "Do you personally choose your team's CRM?" invites a yes. "Who on your team has the final say on the CRM?" doesn't.
2. Keep it to two or three questions. It's a check, not a second survey.
3. Make each one answerable in a line. Open enough to be honest, short enough to type on a phone.
4. Ask about facts, not opinions. Role, responsibility, tools, timing, company size.
5. Order from broad to narrow. Lead with the question most people will pass, so the chat doesn't feel like a test.

Then write down who qualifies in plain words. The agent compares answers against that, so be specific: "Qualifies: makes or signs off the final decision on expense software, and the company has 50 to 500 employees. Doesn't qualify: only uses the software, or works at an agency or consultancy."

The sequence, day by day

Days count from the first step. Messages after the invitation wait until it's accepted.

Notice what the invite asks for: permission to ask a couple of questions, not a commitment to the survey. That's a smaller yes, and it sets up the screener honestly.

What the agent does with each answer

Once someone replies, the reply agent runs the screener:

• They're interested: it asks the screening questions from Knowledge, one per message. If you haven't written any, it hands the chat to you rather than improvising.
• Their answers fit: it tells them they're a fit and sends the survey link with how long it takes and how the reward is paid.
• They don't fit: it thanks them warmly, says this study needs a different profile, and ends the chat. It never sends them the link.
• It can't tell: it asks one short clarifying question. If it's still unclear, the chat comes to you.

The same research rules as every survey template apply on top:

• The link goes out once, and only after they qualify and agree.
• Questions about the reward, payout, who's running the study or confidentiality are answered only from Knowledge; otherwise you get the chat.
• "I've finished, when do I get paid?" always goes to a person.
• A broken link gets a one-line apology and a hand-off.
• A no gets a thank-you and the chat ends.

Follow-ups after they go quiet

People sometimes answer the first screener question and then disappear. Two follow-ups go out: the first 2 days after the conversation stalls, the second 5 days after that: the first picks up where it stopped with one short line, the second is a no-pressure close that mentions your closing date if there is one.

Managing quotas

Quotas are where chat screening pays off most. Put them in Knowledge in the same plain language:

Screening questions:
1. Roughly how many people work at your company?
2. Who has the final say when your team picks expense software?
3. When did your team last review or change that software?

Who qualifies: final decision maker or sign-off on expense software,
company of 50 to 500 people, reviewed or changed it in the last 2 years.

Quotas: 
- Companies of 50 to 200 people: FULL. Thank them and say this group is full.
- Companies of 201 to 500 people: open.

When a cell fills, edit that line. The agent uses the latest Knowledge on the next reply. Anything outside the rules you've written goes to you rather than being guessed.

Keep screening in the survey too

Chat answers are self-reported, the same as form answers. Keep your survey tool's own screener as a second check, and use the same wording so the two agree. Set the survey's webhook to fire only on completed responses, so someone screened out inside the survey isn't thanked for finishing, and write in Knowledge how screen-outs are treated (a smaller reward, or none) in case they ask.

Thank qualified respondents automatically

1. Add ?cid={ContactId} to the survey link set on the campaign.
2. Add a hidden field named cid in your survey tool. Typeform and Tally both read hidden fields from the link.
3. Turn on the webhook under AI Agent → Webhooks & external data and paste its URL into your survey tool or a Zapier step.

When a qualified respondent submits, they get the thank-you straight away, and you can see in the contact list who's finished.

Common mistakes

• Leading questions. If the right answer is obvious, people give it.
• Too many questions. Past three, it feels like unpaid work before the paid work.
• Vague qualification rules. "Senior people in finance" leaves the agent guessing. Spell it out.
• Forgetting to close full quotas. Update Knowledge the day a cell fills.
• Not testing a fail. In the simulator, play someone who doesn't qualify and check they're thanked and never sent the link.

How to launch it in Updately

1. Build the list with a Profile search (titles, industries, company size) or a Sales Navigator list.
2. Create a campaign, open the AI Agent tab, click Start from a template and pick Screen first, then send the survey.
3. Write the screening questions, who qualifies and any quotas in Knowledge, along with the topic, length and reward.
4. Put the survey link with ?cid={ContactId} on the campaign and turn on the webhook.
5. Test a pass, a fail and an unclear answer in the simulator, then launch.

If your study has no strict criteria, the simpler Paid survey invite is a better fit. For one-to-one research calls, see recruiting paid expert interviews. All four research templates are on the campaign templates page, and the LinkedIn survey campaigns guide compares the wider options.

FAQ

What is a screener question?

A short question asked before a survey to check the person fits the study's criteria, such as their role, their responsibility for a decision, or the tools they use. People who don't fit are thanked and not sent the survey.

How many screener questions should I ask on LinkedIn?

Two or three. Enough to confirm the criteria a profile can't show, few enough that it doesn't feel like work.

Will people lie to qualify?

Some will, in chat or in a form. Avoid revealing the qualifying answer, keep the survey's own screener, and look for consistency between the two.

Can I change the criteria mid-study?

Yes. Edit who qualifies or the quotas in Knowledge and the agent follows the new rules from the next reply.