Toolkit
Free market research tools
How results are checked
Named statistical review pending. Selected automated reference cases are published on the validation page.Quality explains
30.6%
of the explained variance in overall satisfaction, ahead of service at 24.9%. Model R² = 0.705.
Share of R² by driver
- Quality30.6%
- Service24.9%
- Price18.4%
- Speed9.6%
- Range9.2%
- Website7.3%
Try any tool on the sample
Research tools, live in your browser
Every tile shows a result computed from the synthetic 600-respondent snack-brand tracker that each tool loads with “Use sample data”.
TURF Analyzer
Find the combination of products, flavors or features that reaches the most respondents, with incremental reach ladders.
Open toolBest 3 flavors
80.3%
reach: sea salt + chili lime + salt vinegar
Incremental reach ladder
Cumulative reach after each added flavor: sea salt 48.7%, chili lime 69.2%, salt vinegar 80.3%, bbq 86.3%, cheddar 89.5%, sour cream 91.7%. Base 600 respondents.
Data Cleaner
Flag speeders, straight-liners, duplicate respondents and incomplete interviews, then export a cleaned file.
Open tool54 / 600 flagged
- Speeders 26
- Straight-liners 38
- Duplicates 4
Key Driver Analysis
Decompose R² across correlated drivers with Shapley values and Johnson's relative weights, plus a quadrant map.
Open toolPriority map · importance × performance
- Fix firstPrice
- MaintainQualityService
- Lower prioritySpeedRange
- Possible overkillWebsite
NPS Analysis
Net Promoter Score with a confidence interval, a two-group significance test, and NPS by segment from your data.
Open tool−14.5
95% CI −20.8 to −8.2
Penalty Analysis
Mean drops and weighted penalties for too-little and too-much JAR responses, with the 20% action threshold.
Open toolLargest penalty
sweetness: too much
- Respondents
- 40.8%
- Liking drop
- 1.65
Segment Builder
Cluster respondents on standardized attitudes, compare solutions by silhouette, and profile each segment.
Open toolk = 2 · silhouette 0.273
Segment sizes: 308, 292 of 600.
Sample Size
Sample size for a target margin of error, finite population correction, design effect, and margin of error from n.
Open tool385
completes for ±5% at 95% confidence (p = 50%)
Label Mapper
Turn numeric codes into readable labels, merge categories, and export a labeled file ready for crosstabs.
Open toolExample mapping
- 1Female
- 2Male
- 3Non-binary
Everything, by stage
The full toolkit
Crosstabs is a set of browser-local market research tools: banner tables with significance letters, survey data cleaning, label mapping, TURF, penalty analysis, relative-importance key drivers, NPS, segmentation, open-end coding and editable report exports. There are no plans or sign-ups, and data files are not uploaded.
Prepare data
Get a respondent-level file into shape before you tabulate it: open it, check quality, and make codes readable.
- Open a data fileImport CSV, TSV, XLSX or SPSS .sav into the browser workspace. The file is parsed on your device.
- Data CleanerFlag speeders, straight-liners, duplicate respondents and incomplete interviews, then export a cleaned file.
- Label MapperTurn numeric codes into readable labels, merge categories, and export a labeled file ready for crosstabs.
- Open a .sav file without SPSSRead SPSS files with variable and value labels, then crosstab and export.
- Supported import formatsEncodings, headers, value labels, size limits and import error messages.
Analyze
Banner tables with significance testing, plus the specialist analyses market researchers run on top of them.
- Banner tables and tab booksRow sets against multibanners with filters, weights, tests and export recipes saved together.
- Significance lettersColumn-proportions tests shown as A/B/C letters on banner tables.
- Multi-select questionsSelect-all-that-apply questions tabulated as % of cases with per-option tests.
- Survey weightingWhat a weight means, how weighted percentages and bases are reported, and the limits on weighted tests.
- TURF AnalyzerFind the combination of products, flavors or features that reaches the most respondents, with incremental reach ladders.
- Penalty AnalysisMean drops and weighted penalties for too-little and too-much JAR responses, with the 20% action threshold.
- Key Driver AnalysisDecompose R² across correlated drivers with Shapley values and Johnson's relative weights, plus a quadrant map.
- Outcome modelingRegularized logistic outcome models with holdout metrics and observed segment evidence in the workspace.
- NPS AnalysisNet Promoter Score with a confidence interval, a two-group significance test, and NPS by segment from your data.
- Segment BuilderCluster respondents on standardized attitudes, compare solutions by silhouette, and profile each segment.
- Open-end codingDefine themes, review assignments and materialize coded variables without sending responses to a hosted model.
- Tracker wave checksDetect and repair variable, type, category and order changes between survey waves.
Plan
Size a study before fieldwork, or check how precise an existing sample is.
Report
Move results into the formats your deliverables already use.
- Export formatsCSV, XLSX, PDF, PNG, SVG, editable PPTX and DOCX: what each contains.
- Editable PPTX and DOCX refreshUpdate generated values in an existing deck or document while keeping supported manual edits.
- Crosstab AtlasWorked example tables with tests, effect sizes and plain-language interpretation.
From first question to next wave
Build on your work. Keep the context.
Make a comparison today, then carry the setup, decisions, and evidence into the next question.
- Ask more without starting over.
- Save a tab book with reusable question sets, audience banners, and filters. Keep table order and analysis choices together so the next run starts with a plan.
- Explore reusable tab books
- See the size of a finding, and what supports it.
- Read counts and percentages alongside statistical tests, effect sizes, and diagnostics. Check sparse cells, missing responses, and sample bases before drawing a conclusion.
- Find the right statistical test
- Keep weights and multiple answers meaningful.
- Use supported, explicitly defined weights and review raking balance and effective-base diagnostics. For multi-select questions, distinguish percentages of people from percentages of responses. Weighted inference requires a supported weight type or approved survey design.
- Weighting guide Multi-select guide
- Compare waves with fewer hidden surprises.
- Compare compatible tables across saved waves. Detect schema drift, review changes to variables and categories, and approve reversible repairs. A matching schema still needs your judgment about question meaning.
- Understand wave comparability
- Pick up where you left off, with a record of why.
- Saved projects retain analysis definitions, revisions, and audit history. Export a project package for reuse; include respondent data explicitly when you need a full-data handoff.
- Explore local projects and portability
- Bring open answers and outcomes into the picture.
- Manually code open-ended responses into reviewed themes and apply them as project variables. Explore binary outcomes with an unweighted logistic model and holdout evidence. Predictions describe associations, not causal effects.
- Open-end coding Outcome modeling
Where your data goes
Files you open in these tools are read and processed in your browser. See local mode for what runs on your device, capacity for workload limits, and MCP for the separate AI-client path, which runs on your machine.