Longitudinal Analysis & Forecast compares your survey responses across time frames you choose, shows how each answer changed, and projects where the results are heading. You get one report with trends, significance checks, AI summaries and a forecast.
Pre-requisites:
A survey with responses in at least two time frames.
Access to Analytics > AI Analytics on a plan that includes AI analysis.
The results date filter set to All Responses, so older responses are included.
How to Set It Up
Open your survey, go to 5. Analytics > AI Analytics and click Conduct Longitudinal Analysis.
Time Frames: pick a Preset (Quarters, Months, Year over Year, Same period each year or Custom, up to 10 time frames). Responses per time frame shows how many responses each one has. Click Continue to Attributes.
Attributes: check the Analysis name, choose the metrics (Responses, View/Drop Counts, Engagement Metrics) and the questions. Questions tagged Personal data start unselected. Click Continue to Forecast Period.
Forecast Period: choose 1, 3, 5 or 10 years and click Generate Longitudinal Analysis. The report is saved under Recent Reports.
The report has three tabs: Trend Analysis (how answers changed across time frames), Projected Analysis (the forecast with a likely range) and Individual Reports (one time frame at a time). Use the % / Count toggle to switch between percentages and response counts.
Example: Employee Satisfaction Survey
An HR team has run the same Employee Satisfaction Survey every year since 2022, collecting about 1,000 responses a year. They set the date filter to All Responses, chose the Year over Year preset with 2022, 2023, 2024 and 2025, kept all 10 questions and picked a 5 year forecast.
Trend Analysis. The AI summary answered their main question: "Employee satisfaction and recognition have improved significantly since 2022." The biggest changes:
Answer | 2022 | 2025 | Trend |
Contributions are always acknowledged | 11.9% | 20.8% | +8.9 pp |
Very satisfied with management communication | 12.8% | 21.6% | +8.8 pp |
Not at all motivated by company initiatives | 22.7% | 14.3% | -8.4 pp |
Average professional development rating (1 to 10) | 5.39 | 6.08 | +0.7 |
Each change was marked Significant (p below 0.001). In the open-text answers, requests for Wellbeing support grew from 19.5% in 2024 to 28.5% in 2025, which showed the team what to focus on next.
Projected Analysis. With four steady years of data, most forecasts had medium confidence. If the trend holds, "very satisfied" with management communication reaches about 34.7% by December 2030 (likely range 31.8% to 37.6%). Work environment satisfaction had low confidence because its trend was uneven.
Individual Reports. Selecting 2025 showed that year alone: 999 responses and an average completion time of 155 seconds.
Terms Used in the Report
Term | Meaning |
pp (percentage points) | The difference between two percentages, for example 11.9% to 20.8% is +8.9 pp. |
Trend | The last time frame minus the first. |
p-value | The chance a change happened by random variation alone. Below 0.05 counts as significant. |
Small sample (under 30) | A time frame has fewer than 30 answers to that question, so treat its change with care. |
+1 year, +2 quarters | Forecast steps counted from the end of your last time frame. |
Confidence | How far to trust a forecast. Two time frames always give low confidence; high needs five or more with a clear trend. |
Downloading and Sharing
Use the download icon to export all the data as CSV, or the tab you are viewing as PDF. Click Publish to create a public link you can copy or open. Charts use your brand kit colours.
Tip: Use at least three time frames of similar length for more dependable forecasts. With only two, treat the projection as a direction, not a prediction.
If you are still running into issues, contact us at [email protected] at any time.