Upload your survey data, choose your banner, and get a complete deck of cross-tabulation tables back in your inbox as a formatted Excel workbook, with an analytical context prompt to help you turn the numbers into a report. Every question tabled against the groups that matter to you, reliable differences flagged, no manual tabulation.
A complete tables deck, delivered as a finished Excel workbook
A light quality read runs on your data before you pay, so you can see what you are working with.
Four steps from your survey export to a finished tables deck in your inbox.
CSV, Excel or SPSS, straight out of your survey platform. No template to fill in, no reformatting.
Point us to your respondent ID, your weight if you have one, and the variables that form your banner, the groups you want every question read against.
Every question tabled against your banner, the reliable differences flagged, and the option to screen the data for quality first.
A formatted Excel tables deck and an analytical context prompt, ready to read, share, and turn into a report.
Four things every table gives you, so the story comes off the page rather than out of a calculator.
Every answer shown as a share of each banner group. Read straight down a column and you are comparing groups like for like, no mental arithmetic.
Each cell indexed against the overall result and heat-shaded, so the groups that sit well above or below the norm catch your eye before you read a single number.
Differences between groups that hold up to testing are flagged, and the marking allows for how many groups were compared. A flagged difference is one you can put in a slide with confidence.
Every column carries its base, so you can see how many people sit behind a figure and judge how much a difference really carries before you build a decision on it.
The tables deck, an AI-ready context prompt to help you write it up, and a VAT-compliant record. Everything you need to read the data, share it, and close the loop on procurement.
Every question in your survey tabled against your banner, heat-shaded, with reliable differences flagged and weighted bases shown. Open it, read it, filter it, drop tables straight into your report.
An AI-ready markdown file that summarises what is in the tables. Paste it into ChatGPT, Claude or Gemini to interrogate the numbers, draft commentary, or start a report or presentation with the full picture already in the conversation.
Depending on your payment path. South African orders include 15% VAT shown at checkout. Orders outside South Africa are zero-rated.
Strong tables come from clean data. So we look at yours first, and we tell you what we find.
A light quality read runs on your file before any money changes hands, so you can see whether your data has issues before you commit to a deck. This is the honest part. We would rather you knew what you were working with up front than found out halfway through a report.
For a small addition, the full screen gives you:
The screen is routing-aware, so respondents who were legitimately skipped past sections are not wrongly flagged as poor quality. You stay in control of exactly what goes into the tables.
One flat price per deck. No setup fees, no per-respondent line items. Add the data-quality screen if you want it, or leave it.
Over 20,000 responses, or need something tabled a particular way? Get in touch.
Look at the deliverable before you upload anything. The tables deck and the context prompt, on sample data.
A complete Excel tables deck built on a sample survey, plus the analytical context prompt that ships with it. Same structure, same heat shading, same significance flagging your own file will produce.
Sample tables deck Sample context promptSamples are being finalised. If a link is not live yet, email greg@knowsis.net and we will send them straight over.
Tables tell you a great deal. They do not tell you everything. Here is where the line sits.
A table can show you that two things go together. It cannot, on its own, tell you that one drives the other. If a group scores high on satisfaction and high on loyalty, the tables put that in front of you clearly. Whether the satisfaction is causing the loyalty, or both are being driven by something else, is a different question, and an important one to keep in mind before you act.
Reading tables well means holding that distinction. It is also exactly where the next analytical steps begin.
Where your tables sit in the wider Knowsis Data Alchemy framework, stated as the genuine next steps an experienced analyst would take on the same data.
The full picture lives on the Data Alchemy platform.
Self-serve in delivery. Practitioner-grounded in method.
Knowsis is an analytics consultancy founded by a researcher who has spent over twenty years working on cross-tabulation, driver analysis, and customer experience analytics across consumer, B2B, financial services, FMCG, and tobacco categories. The standards behind Tables come from that practice.
The analytical standards behind Tables are the same ones in use with our clients today.
JTI · Capitec · FT Longitude · dunnhumby · NielsenIQ · Shoprite · Trade Intelligence · Embark Insight
Your file is used only to build your deck. It is not used as training data and it is not shared with any third party. Your data is not retained beyond the short window needed to answer a support query. Knowsis is POPIA compliant.
Honest answers to the questions most people ask before they upload.
A CSV, Excel (.xlsx) or SPSS (.sav) file straight out of your survey platform. There is no template to fill in and no reformatting to do. Your existing export is fine.
You tell us which column is the respondent ID, which is the weight if you have one, and which variables form your banner.
A banner is the set of groups you want every question read against, the columns that run across the top of each table. Region, age band, customer type, product held, whichever splits carry your story.
Choose the groups a decision will actually turn on. If you are not sure, start with the two or three you would put in a summary slide. You can always table more.
The deck lands in your inbox when it is ready, typically within one working day of your order.
You do not need to keep a tab open or check back. We email you when it is done.
Yes. If your file already carries a weight, nominate the weight variable when you order and every table is produced on the weighted data, with the weighted base shown so you can see how much each figure carries.
If you have no weight, the tables run unweighted. Either way, the base is always visible.
Differences between groups that hold up to testing are flagged in the deck, and the marking allows for how many groups were compared, so a flagged difference is one you can talk about with confidence.
You choose the confidence level when you order.
Yes. Your file is used only to build your deck. It is not used as training data and it is not shared with any third party.
Your data is not retained beyond the short window needed to answer a support query. Knowsis is POPIA compliant.
A light quality read runs on your file before you pay, so you can see whether your data has issues before you commit. That read is included.
The optional full screen scores every respondent for quality, gives you a list of the flagged cases for your own records, and lets you build your deck on all the data or on the cleaned subset. It is routing-aware, so respondents legitimately skipped past sections are not wrongly flagged.
It is an AI-ready markdown file that summarises what is in your tables. Paste it into your preferred AI assistant and you can interrogate the numbers, draft a commentary, or start a report or presentation with the full picture already in the conversation.
It saves you retyping figures by hand, and it keeps the numbers the assistant is working from accurate.
Upload your survey data, choose your banner, and get a complete deck of cross-tabulation tables back as a formatted Excel workbook, with an analytical context prompt to help you write it up.
Build your tablesA light quality read runs on your data before you pay, so you can see what you are working with.
