The shape of the data picks the chart

The problem

Every general-purpose spreadsheet tool ships the same workflow: select a range, open a chart menu, pick a type, discover it was the wrong type, try again. That five-minute detour happens before anyone can look at the actual data, and it happens every single time a new table shows up, because the tool has no idea what shape the table is in until a person tells it.

The people who feel this most are not analysts. They are the ops manager with a weekly export, the marketer with a traffic CSV, the founder with a survey result — people who know exactly what question they want answered and have to fight a chart-type dropdown to ask it. The cost is not the five minutes; it is that most of them give up and read the raw table instead, so the pattern in the data never gets seen.

Why now

Two things make this buildable as a client-side tool rather than a SaaS product. First, real chart-selection logic — detecting a column's type, counting its distinct values, checking whether it parses as a date — is cheap enough to run in a browser on a keystroke, no server and no round trip needed. Second, people have been burned enough by "AI" data tools that oversell inference they don't have; a tool that is visibly rule-based, and says so, is now a credibility advantage rather than a limitation.

The market has also drifted toward no-code and one-shot tools generally, so a product whose entire pitch is "skip the menu" fits how this audience already expects to work.

What it is

A page that takes a table — three bundled examples, or a pasted CSV — reads every column's type and cardinality, and picks the chart form a rule implies: a low-cardinality category next to a number becomes a bar, a date next to a number becomes a line, two numbers become a scatter. The reasoning panel shows its work, in the same terms an analyst would use, so the choice is inspectable rather than asserted.

Suggested views let a visitor look at the same table through a different pair of columns without touching a menu, and the chart re-infers itself each time. Those suggestions are a fixed script, not a conversation, and the interface says so — there is no model behind the questions, only behind which columns they point at.

Who it's for

Someone who already has the table and just wants to see it: an operations lead with a weekly numbers export, a small marketing team with site metrics, a founder with a survey result, none of whom have an analyst on staff or the patience for a charting tool's onboarding.

It is explicitly not for anyone who needs multi-table joins, statistical modeling, or a dashboard that updates on a schedule — that is Tableau's job and it does it better. This tool's entire value is the five minutes it saves on the first look at one table, not a replacement for real analysis.

How it makes money

Free for a single pasted table, because that is the entire hook and gating it would kill the thing that makes people try it. The paid tier is for people who do this often enough to want it saved: connect a live spreadsheet or database so the inferred view updates automatically, export the chart as an embed, and keep a history of tables instead of re-pasting each one.

What is deliberately rejected is a general BI-tool roadmap — joins, filters dashboards, teams. Every one of those turns this into a worse version of a product that already exists and is good at it. The paid tier stays scoped to "the same convenience, on a schedule," not "more power."

Go to market

Launch where people already paste tables into things: a browser extension angle and a "paste your CSV here" link shared directly into ops and marketing-adjacent communities, Slack groups, and subreddits where someone just asked "what chart should I use for this." The tool answers that exact question in one paste, which makes it forward naturally in threads like that.

Second channel is embedding: offer the reasoning panel's rule-based framing as a short explainer piece ("how a computer decides bar vs. line without AI") to data-adjacent newsletters, which reaches exactly the audience that owns a spreadsheet and has never known what to do with it.