Survey results analysis with charts, cross-tabs and coded comments
Your survey closed and now you have a spreadsheet of answers and a pile of free-text comments. Attach the export and your todo.is agent cleans it, counts every question, compares groups, codes the open answers into themes and writes the findings as slides you can present.
The prompt
- Analyze the survey results in the attached file [SURVEY FILE]. The survey was sent to [WHO ANSWERED] and the main question I need answered is [MAIN QUESTION]. Clean the data first: remove empty and duplicate responses and tell me how many you removed. For each question give counts and percentages with a chart. Compare results by [GROUP TO COMPARE]. Code the open-text answers into 5 to 8 themes with counts and 2 example quotes each. Point out results based on fewer than 30 answers. Deliver a PowerPoint of about 10 slides with the key findings first, plus an Excel file with the cleaned data and tables.
What to change
- [SURVEY FILE]: Attach the export (CSV or Excel) from Google Forms, Typeform, SurveyMonkey or similar.
- [WHO ANSWERED]: e.g. "our 1,200 customers", "employees in 3 offices" or "event attendees".
- [MAIN QUESTION]: The decision behind the survey, e.g. "should we add a cheaper plan?".
- [GROUP TO COMPARE]: A column to split by, e.g. "plan type", "department" or "age group".
Example result
- Survey results: customer survey, September
- 1,148 responses received, 46 removed (31 empty, 15 duplicates). 1,102 analyzed.
- Main question: should we add a cheaper plan?
- Key findings
- • 41% of Basic plan users say price is the main reason they might leave, versus 12% of Pro users.
- • 63% would "probably" or "definitely" use a cheaper plan with fewer seats.
- • Satisfaction is high overall: 8.1 out of 10 average, 7.4 for Basic and 8.6 for Pro.
- • Users in their first 3 months are the most price-sensitive group.
- By plan type
- • Basic (n = 702): satisfaction 7.4, would use cheaper plan 71%
- • Pro (n = 364): satisfaction 8.6, would use cheaper plan 48%
- • Team (n = 36): small group, read with care
- Open-text themes (640 comments)
- • Price for small teams · 168 · "We're two people paying for five seats."
- • Mobile app · 121 · "I'd use it daily if the app worked offline."
- • Onboarding · 97 · "Took me a week to find the reports."
- • Integrations · 84
- • Support speed · 59
- • Praise for reports · 111
- What it means
- A cheaper plan for 1 to 2 users looks worth testing, especially for new customers. Price is not the only issue: onboarding and the mobile app come up often.
- Method notes
- • Percentages are of people who answered each question.
- • The Team group has fewer than 30 answers, so its numbers are only a hint.
- • The survey is not random: happier customers may answer more often.
How to do it with todo.is
- Export your survey responses as CSV or Excel and attach the file.
- Copy the prompt, fill in who answered, your main question and the group to compare, then send it in todo.is.
- Your agent cleans the data, builds charts and cross-tabs, codes comments and creates the slides and Excel file.
- Ask for extra cuts like "split by country" or "one slide per theme" until the deck fits your meeting.
Tips for a better result
- State the decision the survey is for. Findings are sharper when the agent knows what you will do with them.
- Keep group sizes in mind. Differences between groups under 30 people are often just noise.
- Ask the agent to show the exact question wording on each slide so nobody misreads a chart.
- Remove personal data like emails from the export if you don't need it for the analysis.
survey results analysis: FAQ
- How do you analyze open-ended survey answers? Read a sample, define 5 to 8 themes, tag every answer with one or more themes, then count them and pick quotes. Your agent does this for all answers and shows the counts so you can check them.
- What is a good survey response rate? It varies a lot. Internal employee surveys often get much higher rates than customer surveys, which commonly land somewhere between 5% and 30%.
- Can it run statistical tests? Yes. Your agent can run tests like chi-square or t-tests in Python and explain what the result means in plain words.
- Which survey tools does this work with? Any tool that exports to CSV or Excel, such as Google Forms, Microsoft Forms, Typeform or SurveyMonkey.
JavaScript is required to use the todo.is app.