A chart plotting paired values as points, one variable per axis. The pattern the points make shows whether, and how, the two are related.
Two columns of numbers, one question: are they related? Paste the data, ask for a scatter plot, and the answer shows up as a pattern, with a trend line fitted on request.
Drag & drop your data
and
600,000+500,000+data professionals
A chart that plots paired values as loose points: one variable on each axis, one dot per observation. The shape the dots make is the finding: a rising band, a falling one, distinct clusters, or no shape at all. It also answers to scatter chart, scatter graph, scatterplot, and, in quality-control circles, scatter diagram. Same chart, five names.
1Add the two columns
Upload a file, paste a table, or connect a sheet. Point at the two numeric columns you want on the axes, then "ad spend vs signups, one point per week" and the plot draws itself.
weekspendsignupsW-04$4,200318W-05$6,800474W-06$5,100402
Ad spend vs signups, one point per week
2Interrogate the pattern
Fit a trend line, color by category, log-scale a skewed axis, label the outliers that deserve names.
Select chart typeScatter
Line of best fitColor palette
3Export and share
PNG, SVG, or PDF for the report, or an interactive version where every point answers on hover.
PNG
This is what it looks like in the notebook: one prompt in, an explained, editable plot out.
Spend efficiency
Ask AI
Ad spend vs signups, one point per week, add a line of best fitSpend and signups move together — r = 0.87 across 24 weeks, roughly 62 signups per extra $1,000. Two weeks sit well below the line; both ran during the pricing-page outage.
Weekly ad spend vs signupsWeekBest fit
Four shapes, four different answers to the same question.
Positive correlation
Points climbing from lower left to upper right: as one variable rises, so does the other. The tighter the band, the stronger the link.
Negative correlation
The downhill version: one goes up, the other comes down. Price against demand, age against top speed.
No correlation
A shapeless cloud. An unglamorous but useful answer that ends arguments about relationships that were never there.
Clusters and outliers
Sometimes the finding isn’t a line but a grouping: two distinct blobs, or one point far from everything. Scatter plots are how those get caught.
Understands the data
Ask for a line of best fit, a correlation coefficient, or log axes and they appear. A basic point-drawing tool leaves those parts for you to work out yourself.
Stays connected
Points stream in from CSV, Excel, Google Sheets, or a live database, so the pattern updates as new observations land.
Built for sharing
Publish an interactive plot where hovering any point reveals its values, embed it in docs and sites, or export clean stills for slides.
The same AI generator handles every common chart — pick the one that fits your data.
Paste the two columns you've been arguing about and let the pattern settle it.