Know which results you can stand behind.
Drop in your results spreadsheet. Every check runs at once, and the summary sorts your findings into what's safe to say, what needs a caveat, and what a quick read of the totals would get wrong.
One PIN, both apps. This Analyzer and the Survey Validator, six months for $49.99 USD.
Excel or CSV, one row per person, from Microsoft Forms, Qualtrics, SurveyMonkey and most survey tools.
You decide who counts, before you see the results
The Analyzer flags answers that look rushed or careless and asks whether to keep them before it shows a single result, so the choice can't be bent to suit the numbers.
Change your mind later and the summary, the PDF and the written report all say so, and what you chose first.
What you can say, and how to say it
The summary sorts every finding into three groups. It tells you when a high score is the most it could be rather than a real result, gives you the words to report it, and lists what a spreadsheet or your survey tool's own report would get wrong.
Every figure comes from calculation, not AI. A record of how each was made goes with the summary PDF and the written report.
Every check, at once
Each one has its own page, linked from the summary, and each page says what the figures do and don't support.
Response quality
Rushed and straight-line answers, missing data, and how many people the results rest on.
Question by question
Each answer scale read on its own terms, never averaged with a different one.
When the questions shaped the answers
Agree-by-default wording, unbalanced scales and similar effects, with how to report the figure honestly.
Scale reliability
Whether the questions meant to measure one thing actually hang together.
Hidden groupings
Which questions move together, so you can see the themes your people actually answered.
Group differences
Compares teams, sites or roles, and holds back groups too small to read safely.
Written comments
Sentiment and topics, and a check of the comments against the people who gave each rating.
Ask in plain words
Ask about your results, including what a sceptical manager would challenge, and get answers grounded in the figures.
Report and slides
A summary PDF, a written report and presentation slides, each recording how the figures were made.
What uses AI, and what doesn't
Every figure is calculated, never generated. AI only writes about the figures, only when you ask, and only after you've agreed to it once per session.
- Every check, chart and figure above
- The summary and the summary PDF
- Sentiment and topics in comments
- The record of how each figure was made
- A plain-words reading of a page
- The written report and the slides
- Naming themes and groupings
- Explaining what comments say about a low score
- Answers to your questions
In our own use, a full run with page readings, a written report and slides costs about 30 to 90¢ of the $12.
What the AI sees
Computed results: averages, reliability figures, counts, topic word lists, and your question names. Never your uploaded table. The one feature that sends comments themselves, explaining a low score, asks every time and takes out emails and phone numbers first. Names people wrote in comments are not removed.
What's kept
Your file and results last only for your session: download what you need. Questions you ask are saved under your PIN so you can look back at them, and you can delete them. Privacy statement ›
Results in already?
One PIN, six months from your first sign-in, both apps. Analyse as many surveys as you need, with $12 of AI included.