Sopact is a technology based social enterprise committed to helping organizations measure impact by directly involving their stakeholders.
Copyright 2015-2026 © sopact. All rights reserved.
A survey is rarely one questionnaire — it's a series of waves you must join on one person. Survey intelligence connects every wave on a single identity and reads open-ended answers alongside the numbers the moment they arrive, so per-person change is provable and the narrative is cited.
A survey is rarely one questionnaire — it's a series of waves the same people move through over time: a baseline, a mid-point, an exit, a follow-up a year later. Most survey tools store each wave as a separate export; they can tell you what a group answered this month, never how one person changed unless someone joins the files by hand. Survey intelligence reads every wave as one connected record on a single identity, the moment each response lands — so did this person actually change, and why? becomes a question you can answer with their own words.
Key takeaways
Survey intelligence means treating every wave a respondent answers — closed questions, open text, across baseline, mid, and follow-up — as one connected record on a persistent identity, and reading it the moment it arrives instead of at analysis time.
If you run a longitudinal or mixed-methods study, you already collect at every wave. But each wave lands as its own spreadsheet, open-ended answers pile up unread, and by analysis nobody can line up one person's baseline against their follow-up — or say why the number moved. Teams report the averages they can compute and leave the open text in an appendix. The question funders now ask has moved from what did the group score to did individuals change, and can you show why — and disconnected exports cannot answer it.
A survey tool — SurveyMonkey, Qualtrics, Google Forms — collects responses and charts the averages. Survey intelligence reads and connects them. Three shifts make the difference:
None of this requires a new survey platform — survey intelligence reads what your existing forms collect and connects it on the way in.
Forget the one-off-poll connotation for a moment. In this course, a survey is any repeated measurement of the same people over time — whoever they are. A training cohort surveyed at intake, exit, and six months out. A youth program tracking the same students across years. A panel study following households through a policy change. A patient-reported-outcome series across a course of care. Each is many waves that must connect on one identity to mean anything.
Honest boundaries first, because the fastest route to disappointment is applying a good approach to the wrong problem. A strong fit shares three traits: the same people answer more than once, much of the signal is in open text, and someone is owed proof that individuals changed.
| Strong fit | Why |
|---|---|
| M&E & impact evaluation | Outcomes owed to funders; change must be shown per person, not just on average |
| Longitudinal & panel studies | The same respondents across waves — the join on one identity is the whole point |
| Pre / post program evaluation | Baseline vs follow-up per participant, with the reasons in the open text |
| Multi-language & multi-country surveys | Open text cleaned and themed across languages on one record |
| Mixed-methods research | Numbers and narrative read together, every theme cited to source |
| Grant-funded outcome measurement | A cited narrative the funder can trust falls out of the analysis |
| Not the right fit | Why |
|---|---|
| One-off satisfaction polls | A single wave with no identity to join — a survey tool is enough |
| Anonymous market research | No persistent respondent to track change on |
| Real-time operational feedback | Transactional signals for ops, not an outcomes journey |
| Voting & census enumeration | Count-once exercises with no per-person change to measure |
The rule of thumb: if the same people answer more than once and their words carry the evidence, survey intelligence fits.
Throughout this course we follow one worked example — an M&E team running a longitudinal study from baseline to funder report — because it exercises every phase most readers share: design on one identity, read quant and qual together, and write the cited narrative.
Every wave joined on one identity — numbers and open text read together, each theme cited to source.
A persistent ID and a wave design so baseline, mid, and follow-up line up as one person's change — not four disconnected exports.
FoundationOpen text cleaned and themed on arrival and connected to the numbers, every theme tied to a respondent's exact words.
★ Flagship · AnalyzeOne funder-ready narrative where every claim links to the response behind it — the report becomes a query over evidence.
CommunicateSurvey intelligence begins with one decision made before the first wave goes out: a persistent identifier that follows each respondent across every wave, and a wave design that asks the same core measures each time. Get this right and per-person change is provable from the second wave onward; the cited narrative is possible because every answer traces back to a person. The design chapter sets it up.
Take one survey you run more than once and decide how you will recognize the same respondent across waves — the ID, not the email, not the name. That single choice is what turns four spreadsheets into one person's story.
M&E and evaluation leads, researchers running longitudinal or mixed-methods studies, and program teams who must prove individual change — with the reasons, in respondents' own words — to a funder or a board.
Frequently asked questions
Joining every survey wave on one identity and reading open text alongside the numbers on arrival — so per-person change is provable and the narrative is cited to respondents' words.
Those collect responses and chart averages; survey intelligence connects waves on one respondent, themes the open text, and ties every finding to source.
No. It reads what your existing forms collect and connects it on the way in.
They're cleaned and themed the moment they land — across languages — and connected to the numbers, each theme cited to the respondent's exact words.
Yes — a persistent ID joins every wave, so baseline, exit, and multi-year follow-up line up as one record.
Survey Intelligence is the canonical home for longitudinal and mixed-methods survey analysis; Case Intelligence uses these methods inside a broader stakeholder journey and links to them.
Next: Clean Open-Ended Responses at the Source → · or Try the Survey Intelligence demo in Sopact Sense →
Open Sopact Sense, paste your program description, and put it to work.
Try in Sopact