A respondent writes “went well” in a sprint retro field. That's it. Two words. You know nothing more than you did before sending the form. The question was fine. The answer is empty.
Qualitative researchers solved this problem decades ago. In a live interview, when a participant gives a thin answer, the interviewer doesn't move on. They probe: a targeted follow-up designed to pull out the specifics the participant already knows but didn't bother to say. Patton's Qualitative Research & Evaluation Methods (2002) catalogs the full taxonomy. Willis's Cognitive Interviewing (2004) adapted the technique for survey testing. Miller and Rollnick's Motivational Interviewing (2012) refined it further for contexts where the respondent is reluctant or disengaged.
Forms don't have interviewers. But they can have probes. Here are seven patterns, each with a name, a trigger, and an example of the answer it produces.
Probe taxonomy · 7 follow-up patterns
1. The elaboration probe
Pattern: “Can you say more about [X]?”
Patton (2002) calls this the most basic and most effective probe. It works because it doesn't redirect — it simply asks the respondent to continue on the same thread. The underlying psychology is straightforward: most people stop talking before they run out of things to say. They stop because the form didn't signal that more was wanted.
Thin answer: “Good sprint overall.”
After probe: “Shipped the auth migration with zero rollbacks. Pair programming on the API layer saved us two days of review cycles.”
Configure this as the default probe on any open-ended field. It's the right choice when you have no reason to pick a more specific pattern.
2. The clarification probe
Pattern: “When you say [X], do you mean [A] or [B]?”
Willis (2004) developed this for cognitive interviews — testing whether respondents and question authors share the same understanding of a word. In form context, it fires when the answer uses vague terms. “The process was confusing” becomes “When you say confusing, do you mean the steps were unclear, or that you didn't know where to start?”
Thin answer: “The onboarding was confusing.”
After probe: “I didn't know which fields were required until I tried to submit. The error messages showed up at the top, not next to the field.”
Use this on fields where you need to act on the response. Vague feedback is hard to prioritize. Clarified feedback tells you exactly what to fix.
3. The contrast probe
Pattern: “How does this compare to [previous experience]?”
Patton (2002) notes that comparison forces specificity. A respondent who says “this was fine” can't say “same as before” without defining what “before” looked like. The probe creates a reference point.
Thin answer: “The new dashboard is okay.”
After probe: “Faster than the old one. I can find the response summary without clicking through three tabs. But I miss the export button that used to be on the main screen.”
Use this after launches, redesigns, or process changes — any moment where a “compared to what?” makes the answer richer.
4. The example probe
Pattern: “Can you give a specific instance?”
Miller and Rollnick (2012) call this “asking for the story.” Abstract claims (“communication could be better”) collapse into actionable detail when you ask for one concrete moment. The respondent shifts from opinion mode to narrative mode, and narratives carry specifics that opinions don't.
Thin answer: “Communication could be better.”
After probe: “Last Thursday, the API spec changed after I'd already built the integration. Found out from a commit message, not from anyone telling me.”
Pair this with questions about team dynamics, process friction, or user experience — anywhere an abstract complaint hides a concrete event.
Elaboration probe · before & after
No probe
With probe
5. The consequence probe
Pattern: “What happened as a result?”
This is Patton's (2002) technique for uncovering impact. A respondent mentions an event but stops before explaining why it mattered. The consequence probe extends the causal chain. It turns “staging was down” into a story about three engineers blocked for half a day.
Thin answer: “Staging was down on Tuesday.”
After probe: “Three engineers were blocked for four hours. We missed the Wednesday QA window and had to push the release to Friday.”
Use this on incident reports, risk assessments, or retrospective fields — anywhere the downstream impact matters more than the event itself.
6. The feeling probe
Pattern: “How did that land for you?”
Miller and Rollnick (2012) describe this as an “affect reflection” — a probe that asks about emotional response without leading toward a particular emotion. It works because factual accounts often omit the human reaction. The respondent tells you what happened but not how it felt, and the feeling is often the signal your team lead or manager actually needs.
Thin answer: “The reorg moved me to a new team.”
After probe: “Honestly frustrating. I'd spent two months building context on the payments system and now I'm starting over on a codebase I've never seen.”
Use this for engagement surveys, manager check-ins, and exit interviews. Save it for fields where the emotional dimension matters. On a technical retro, the consequence probe is usually more useful.
7. The quantification probe
Pattern: “Roughly how many times? How long did it take?”
Willis (2004) found that respondents default to qualitative language even when they have quantitative knowledge. “It took a while” might mean 20 minutes or 3 days. The quantification probe asks for the number, and most respondents can provide one — they just didn't think the form wanted it.
Thin answer: “Deployments are slow.”
After probe: “Average deploy takes about 45 minutes. We deploy twice a day, so that's 90 minutes of waiting. Last month it was closer to 30.”
Use this on fields about frequency, duration, or scale. Numbers make reports concrete. “Deployments take 45 minutes each, twice daily” is a sentence that gets budget allocated. “Deployments are slow” is a sentence that gets forgotten.
From taxonomy to form field
Each of these seven patterns maps to a probe you can configure on a Pluck form field. You pick the field, choose a probe pattern, and optionally customize the wording. When a respondent submits a thin answer — short, vague, or abstract — the probe fires as a follow-up question. The respondent sees it immediately and can add detail while the context is still fresh.
You don't need all seven on one form. Start with elaboration as a default, add clarification on your most important field, and quantification on anything where you need numbers. Three probes cover most use cases. Add the rest when a specific field keeps producing answers you can't act on.
The best form is still the one with well-written questions. Probes don't fix bad wording. They catch the gaps that good wording leaves behind. Write the question to set the frame. Let the probe handle the depth.