Surveys

Plan — Creating surveys, asking questions, and reading results are available on every plan. Three things need Startup or above: targeting a survey by respondent traits, exporting responses as CSV, and the Trends section on the results page. Trait rules you already saved keep filtering respondents, and you can always remove them. Upgrade at accounts.frugalsaas.com/billing.

Surveys are targeted in-app microsurveys that measure how your product and your releases actually land. They're the fourth module beside boards, roadmap, and changelog, and they close the loop: this is where heedback measures. Responses are unlimited on every plan — no per-response metering.

Survey types

Start from a template or build your own:

• NPS — Net Promoter Score
• CSAT — customer satisfaction
• CES — customer effort
• PMF — product–market fit ("how would you feel if you could no longer use this?")
• Custom — multi-step surveys you assemble yourself

Question kinds include scale/rating, multiple choice, emoji, and open text.

Delivering a survey

Two delivery channels, both built on the same widget identity so responses are tied to a known respondent:

• In-app widget — show the survey as a slideout or modal to users on your site, with show-once and reshow throttling so it never nags.
• Shareable link — send a tokenized /s/… link (email, DM, etc.). Each invitation is per-recipient and attributable, and can expire.

Targeting

Decide who sees an in-app survey by URL/path (e.g. only on /settings) and by respondent traits you pass from your app. Only matching users are shown the survey.

Reading results

Each survey has a results view that ties every answer back to the respondent, so you can see not just the score but who said it and follow up. Because responses share identity with your boards, you can connect what someone asked for with how they rated it once it shipped.

Trends

(Startup and above.) On an NPS, CSAT, or CES survey, the results page shows a Trends section: your score over time, the number of responses behind each point, and how the promoter / passive / detractor mix shifted.

• Choose weekly or monthly buckets, and a date range (30 days to all time). A very wide range switches to monthly automatically so a single quiet week doesn't read as a swing.
• Periods with fewer than five responses are faded — the score is there, but don't over-read it.
• Filter by a respondent trait (for example plan = pro) to see the trend for just that segment. The section shows how many responses matched.

The trend is read from the survey's first rating or 0–10 scale question. If a survey has none yet, add one and the section appears. The most recent point is the current, still-filling period — it firms up as responses come in.

Where it fits

Ship something from your changelog, then survey the users who requested it — the measurement half of the boards → roadmap → changelog → surveys loop.