Skip to main content

Enterprise MR tools, without the enterprise contract.

The banner tables, weighting, TURF, and on-device AI coding agencies run on $15k/year platforms — self-serve in your browser, with your survey data never leaving your device. Passes security review in minutes, not months.

No IT approval. No DPA. No procurement queue.

Banner tables with sig letters. Export to Excel in seconds.

Significance testing. No SPSS license. No install.

Open-end coding. No PII leaves your machine.

1,208 tests passing  ·  Zero data transmitted  ·  No install required

Try it. Right now. No signup.

Drag variables to build a live banner table from real survey data — computed in your browser.

Demo: Consumer Electronics Brand Tracker, Q1 2026  ·  n=247
Variables
Drag to a zone below, or press Enter to select then tab to a zone
Gender
Age Group
Region
Brand Awareness
Purchase Intent
Satisfaction
NPS Category
Segment
Stub (rows)
Drop variable here
Banner (columns)
Drop variable here
Zero-upload · Data never transmitted
Stats validated against SPSS 29 output
No DPA or data processing agreement needed
Built by PMR experts
Used by agency analysts and consultants
Zero-upload · Data never transmitted
Stats validated against SPSS 29 output
No DPA or data processing agreement needed
Built by PMR experts
Used by agency analysts and consultants

Everything you need. Nothing you don't.

Browser-native tools that replace five-figure software licenses.

For Insight Leads
AI Open-End Coder. No PII to OpenAI.
Run MiniLM sentence embeddings 100% in-browser. Theme open-ended responses instantly without sending respondent data to any third-party API.
Performance
50k
rows per dataset, handled locally
Local-first WebAssembly engine. No server round-trips.
For Agency Analysts
Instant Banner Tables. Stop fighting Excel.
MR-standard crosstabs with significance letters (Holm step-down) in seconds. Export directly to merged-header Excel formats your clients expect.
BrandGen Z (A)Millennial (B)Gen X (C)Total
ABC
Brand A42%B31%29%34%
Brand B18%29%A33%A27%
Brand C28%33%24%28%
Don't know12%7%14%A11%
Brand A42%B
No DPA needed
Data never leaves your browser — verified by architecture, not policy.
For Freelancers
WeightWizard & SurveyStat — no $3,000 license
Rim weighting, automated significance testing (Bonferroni, Holm, FDR), and k-means clustering — self-serve, cancel anytime.
Efficiency84.2%

The full DP workflow. Nothing left out.

From raw CSV to weighted, significance-tested tables — without touching a desktop app.

01
Clean & Prep

Start with clean data.

Data Cleaner FreeLabel Mapper FreeWeightWizard Standard

Flag speeders, straight-liners, and AI-generated responses. Remap variable labels. Apply rim weighting to match population targets. All before you run a single crosstab.

IDTime(s)StraightlineStatus
10492420NoValid
1049312YesFlagged
Speeders found14%

02
BrandGen Z (A)Millennial (B)Gen X (C)Total
ABC
Brand A42%B31%29%34%
Brand B18%29%A33%A27%
Brand C28%33%24%28%
Don't know12%7%14%A11%
Brand A42%B
Crosstab

Banner tables that look like you made them in Q.

Banner Table StandardSurveyStat Standard

Drag-and-drop banner builder. Column percentages, significance letters (A/B/C), net rows, means. Bonferroni, Holm, and FDR corrections. Excel export in MR format — merged headers, sig letters in-cell, ready for a client deck.


03
Analyse & Segment

Go deeper without switching tools.

Key Driver Analysis ProfessionalTURF ProfessionalPenalty Analysis Professional

Identify what drives satisfaction. Optimise claim portfolios with TURF. Find respondent segments with k-means clustering — the segment column writes back to your dataset and appears as a banner break automatically.

Segment Profile · k=3
Brand LoyalistValue SeekerOcc. Buyer
Satisfaction90%30%50%
Awareness85%40%30%
Price agree20%88%55%
Loyalty95%25%35%
Segment column writes back to dataset → use as banner break

04
Open-End Coder · 247 verbatims
Price / value for moneyHIGH
34% of responses
Product qualityHIGH
28% of responses
Customer serviceMED
18% of responses
Delivery speedMED
12% of responses
Returns & refundsLOW
8% of responses
Embeddings computed locally · No verbatim ever uploaded
Code & Export

Code open ends without uploading a single verbatim.

Open-End Coder StandardExcel export StandardSPSS export Professional

AI thematic coding runs entirely in your browser using on-device sentence embeddings. Every category comes with a confidence score — so you can defend every classification to a client. Export to Excel or SPSS .sav format.


No upload. That's not a feature — it's the architecture.

Every other browser-based analytics tool sends your data to their servers. That's why they need a DPA. That's why your IT team blocks them.

Krosstabs runs entirely in your browser using Web Workers and IndexedDB. Raw data is never transmitted. Not even to us. The only thing that reaches our servers is your account email.

Don't trust a diagram. Verify it yourself: open your browser's Network tab, load a dataset, run an analysis. Count the outbound requests that contain survey data. The answer is zero.

Read our security architecture
YOUR BROWSERAll computation happens here.Raw data never crosses this boundary.CSV DataYour local filenever uploadedWeb WorkersIsolated computethreadIndexedDBLocal encryptedstoragemetadata only(account email)Supabaseauth & metadata only

Built for analysts who know the alternatives.

vs. Excel

Excel can't do significance testing, rim weighting, or banner tables with sig letters. You know this — that's why you're here.

See Banner Table
vs. Q Research / SPSS

Q and SPSS are desktop apps that require a license, an install, and IT clearance. Krosstabs runs in your browser in 3 seconds. Same output quality. Different architecture.

See honest comparison
vs. Displayr

Displayr requires data upload — meaning a DPA, an IT review, and a 3-week wait. Krosstabs processes everything locally. Your compliance team can verify this in the Network tab.

See security

What the output actually looks like

Banner table with significance testing — the same view your clients will see. Toggle between unweighted and raking-weighted results.

krosstabs.com/dashboard/tools/survey-stat
Preferred coffee type × AgeLive demo
▲ Significantly higher▼ Significantly lower95% confidence · Welch t-test
Total18–3435–5455+
Espresso32%41%34%19%
Filter28%21%31%35%
Instant18%11%17%30%
Cold Brew14%20%12%9%
No preference8%7%6%7%
Base (n)800298307195

vs. heavyweight workflows

An honest comparison. We win on speed and privacy; SPSS and Q win on breadth. Know what you're buying — and what you're leaving behind.

KrosstabsSPSS / Q ResearchDisplayr
Data leaves your device?No — local-firstYes (cloud/DPA needed)Yes (cloud/DPA needed)
IT approval required?NoYes — months of reviewYes
Setup timeInstant (browser)Days (desktop install)Hours
Pricing$69/mo, cancel anytime~$3k–$10k/year (locked in)~$5k+/year
MR-native outputsBanner tables, JAR, TURFBanner tables, scriptingBanner tables, charts
Learning curveNo-code UIHigh (syntax/scripting)Medium
DPA / data agreementNot requiredRequiredRequired
Max dataset sizeUp to 50,000 rowsMillions of rowsLarge
Scripting / automationNo-code onlyFull syntax & macrosSome scripting
Team seats / collaborationSingle-userMulti-seatMulti-seat
Full comparison: Krosstabs vs. Q Research
1208
tests passing
0
bytes of survey data uploaded
(it stays in your browser — check the Network tab)
<2s
avg computation time
(20,000 rows, rim weighting)

Built by PMR experts, for market researchers. Not by a design agency guessing what analysts need.

Local is blazingly fast

Testing a 50,000-row tracker. Krosstabs crunches it mathematically in the browser while traditional cloud SaaS is still waiting for your file upload to finish.

Enterprise Cloud SaaS
~15s
Krosstabs (Local)
~0.4s

No black boxes.

Every formula is documented. Every result is verifiable against SPSS 29.

See the reproducible SPSS-29 benchmark
Weighting
Raking / IPF
w_ij = w_ij × (T_j / Σ w_ij)
Learn how it works
Significance
Chi-Square + Holm/FDR
χ² = Σ (O − E)² / E
Learn how it works
Segmentation
K-Means Clustering
argmin Σ ‖xᵢ − μₖ‖²
Learn how it works
Key Drivers
Ridge Regression R²
β = (XᵀX + λI)⁻¹ Xᵀy
Learn how it works
See all formulas and benchmarks

Start free. Upgrade when the project demands it.

Academic / QC

Sandbox

$0

For academics, fieldwork QC, and anyone exploring. Free forever. 5k rows.

Recommended
Freelancer / Solo Consultant

Standard

$69/mo

Full MR workflow: weighting, banner tables, sig testing. 20k rows.

Agency / Brand Tracker

Professional

$149/mo

Tracker Harmonizer, k-means clustering, and Key Driver Analysis. 50k rows.

No lock-in. No auto-renewal traps. Cancel from your dashboard in one click.

See full plan comparison

Questions we get asked

Where is the data stored?
On your computer. Specifically, in your browser's IndexedDB. If you clear your cache, it's gone. We don't have a copy.
Can I handle a 50k row tracker?
Yes. The Professional plan supports up to 50,000 rows. It runs surprisingly fast because it's all local.
Is this just a wrapper for R?
No. We wrote the statistical engine from scratch in TypeScript and WebAssembly to run natively in the browser.
Do I need to talk to sales?
No. You can sign up, subscribe, and cancel entirely self-serve.

Prove it works.

Don't take our word for it. Load a sample dataset and check the numbers against your current tool.