What NPS survey software should do beyond scoring: read the open-ended why on arrival so the number arrives as ranked, cited drivers.
NPS survey software should send the 0-to-10 recommend question, calculate the score, and — the part that separates a tool from a scoreboard — read the open-ended “why” on arrival so the number arrives with its reasons. Most NPS tools automate the score and the trend and reduce the comments to a word cloud, which is the half that cannot be acted on. The score is easy to produce; the reasons are the reason to run NPS at all.
The frustration teams describe is a dashboard that trends a number they cannot move: “our NPS software tells us the score dropped and gives us a word cloud, and we still have to read hundreds of comments by hand to find out why.” Software that scores but does not read leaves the actionable part of NPS as manual labor.
Key takeaways
Sopact keeps the rating, open-ended explanation, respondent identity, touchpoint, and later response on one record. Teams can see what drives promoters and detractors and whether the relationship improves after action.
The Net Promoter question was always a pairing — a number and a “why” — and NPS software that automates the number while flattening the why into a word cloud has kept the easy half and discarded the useful one. A word cloud shows that “support” is frequent; it cannot show that support is failing new customers on onboarding while delighting long-tenured ones, because it pools every comment and reads none. So the software produces a trend line and leaves the reasons to whoever has time to read hundreds of comments, which is no one.
Software worth the name reads the why. Sopact calls the record it reads on the connected feedback history: every NPS comment coded against your codebook on arrival, themed and cited, and kept on the same respondent as the score and every later wave. The number arrives as ranked, quoted drivers, not a cloud, which is the mechanics the NPS verbatim analysis page walks through.
NPS software moved through three eras. First, an email survey and a spreadsheet where the score was tallied by hand. Then the NPS dashboard — Delighted, AskNicely, SurveyMonkey and peers — which automated the score, the trend, the segments, and a word cloud, effortless everywhere except the reasons. The current era reads the verbatims into ranked drivers on arrival, tied to the score and the respondent.
a practical buying check that separates the eras: ask the software to explain a two-point drop with the ranked drivers behind it, cited to comments, split by the segment where the drop happened. A dashboard can show the drop and a word cloud; it cannot show the ranked, segmented, cited reason. If the explanation is a word cloud, the software is scoring the number, not diagnosing it.
Good NPS software resolves a tension most tools force on you: to keep the survey short and still get rich reasons. The temptation, when the open-ended answer is unreadable at scale, is to add closed-ended driver questions, which lengthens the survey and lowers response. Software that reads the single open-ended “why” on arrival removes that temptation, because one readable open question outperforms a battery of closed ones and keeps the survey to the two questions NPS should be.
Reading also has to respect the scale’s three segments, because a detractor’s “support” and a promoter’s “support” mean opposite things. Theming detractors, passives, and promoters separately is what turns comments into drivers, and it is exactly what a pooled word cloud cannot do — the same discipline the broader feedback tools apply.
Bring last quarter’s NPS comments to the demo and ask the software to turn them into ranked drivers by segment, cited to source, on a continuous feedback history — not a word cloud. A tool that scores and trends but leaves the reasons unread is a scoreboard. Evaluating on your own comments rather than a clean demo exposes whether the software reads or just counts.
The output that matters is a score that arrives explained: ranked detractor drivers quoted, movement between waves attributable to reasons, and the same respondents followable over time. Because Sopact reads the why on arrival on the connected feedback history, that is what the software produces natively, feeding the CSAT and broader feedback reads on the same record.
NPS scoreboard software trends a number and clouds the comments; software that reads turns the why into ranked drivers. The difference is whether the open-ended answer is read on arrival.
| Requirement | Scoreboard | Reads the why (connected feedback history) |
|---|---|---|
| Calculates the score? | Yes | Yes |
| Reads the comment? | No: a word cloud | Themed into ranked drivers, cited |
| Segments the read? | No: pools all comments | Detractors, passives, promoters apart |
| Follows the respondent? | No: each wave its own pool | Yes: one continuous record |
The reading in depth is NPS verbatim analysis; the sister metric tool is CSAT survey software.
A feedback score is a lagging summary; the reason behind it is in the comment, and only if someone reads the comment while it still matters. That is the premise of the Loop, Sopact’s method for continuous intelligence: collect clean at the source, analyze the moment a response arrives, improve while there is still time to act. A detractor’s reason is themed the day it lands and routed to someone who can respond, not filed for a quarterly readout.
The Loop is also what makes a feedback finding defensible: every driver and theme traces to the exact response it came from, the standard detailed in Loop traceability, so a reported reason is backed by the words, not a hunch.
One method, three moves that never stop
Then the cycle runs again, a little sharper each time. Read the method: the Loop methodology →
Use one real NPS cycle with promoters, passives, detractors, open-ended reasons, meaningful segments, a repeated wave, missing responses, and a relationship decision.
Feedback teams should update the NPS question, follow-up, segments, alert rules, and review status without rebuilding reports.
How to test it
Each score and explanation should stay connected to the correct respondent, account, program, or relationship where consent permits.
How to test it
The system should read every promoter, passive, detractor, and open-ended explanation.
How to test it
Teams should see recommendation change across waves while survey timing, sample, and relationship context remain clear.
How to test it
Open-ended responses should explain promoter and detractor drivers while exact passages remain inspectable.
How to test it
Account notes, support conversations, interviews, and research files should retain source context.
How to test it
A plain-language NPS question should show filters, denominator, included records, and cited comments.
How to test it
A reviewer should reproduce the NPS score and one reported driver.
How to test it
Use one real NPS wave with the rating, verbatim comment, respondent key, segment, touchpoint, and date. The test is whether the platform helps a team decide what to do next, not whether it draws the familiar chart.
Send the 0-to-10 recommend question, calculate the score, and read the open-ended why on arrival so the number arrives with ranked, cited reasons. Most tools score and trend and cloud the comments. Sopact reads every NPS comment on the connected feedback history, so the score is diagnosed, not just displayed.
Because it counts words rather than reasons and pools detractors with promoters, so the same word carries opposite meanings and no driver is actionable. Sopact reads each comment into ranked drivers by segment on the connected feedback history, which a word cloud cannot do.
Bring last quarter’s comments and ask the software to turn them into ranked drivers by segment, cited to source, on a continuous feedback history. A tool that scores but leaves the reasons unread is a scoreboard. Sopact reads the why natively on the connected feedback history.
Yes — without a stable respondent identity you cannot see who moved from detractor to promoter or follow a relationship. Sopact keeps every wave on one continuous record, so movement is attributable and the reasons are followable over time.
The modern kind can: it themes each comment against a codebook on arrival, cites the source, and separates the segments. Older dashboards offer a word cloud. Sopact reads on arrival, so explaining a score move does not require reading hundreds of comments by hand.
Two: the rating and one open-ended why, because software that reads the open answer makes a closed-driver battery unnecessary. Sopact reads the single why into ranked drivers, so the survey stays short and still explains the score.
A dashboard trends the score and clouds the comments; Sopact reads every comment into ranked, cited drivers by segment and keeps respondents on the connected feedback history across waves. So the score arrives explained and movement is attributable, rather than displayed and unexplained.
Next: read the comments in depth on NPS verbatim analysis, or the sister metric on CSAT survey software.
