Transform complexity into clarity.

Native to your dashboard Add it to a worksheet like any Tableau viz. It follows your filters, sorts, and dashboard actions, and refreshes live as the data changes. Nothing to export, connect, or configure.

Drag a dimension to Category and a measure to Value, and the rings render and scale on their own. Everything runs inside Tableau, on your machine: your data never leaves the browser. Free to use, on Desktop, Server, and Cloud.

Which month is the busy one, and is that month getting busier. A line by date answers neither. Here the level between panels answers the first and the trend inside each panel answers the second. On four years of Superstore, November averages 88,167 against February at 14,938, and every month is still visibly growing year on year underneath that.

Drag two to ten measures onto one shelf and the dimension that defines an observation onto Level of Detail. On Superstore at sub-category and region, sixty-eight observations, discounting correlates plus 0.30 with sales and minus 0.29 with profit: it sells more and earns less, which is the sort of thing you find in ten seconds here and never find in a bar chart.

A threshold typed once and left behind is how this chart goes wrong: the data moves and the line stays where it was one afternoon. The default is the median of what is on screen, recomputed every time, with the average, the mid-range, zero and fixed values as alternatives. Whichever rule drew them, the chart prints where its lines came from and whether they move.

A second bar beside the first turns the comparison into a race for length and the value stops being readable. The plan is a needle across the actual bar instead, so how far the bar reaches still means the amount, and the gap gets a panel of its own measured from zero, because a gap is a quantity in its own right.

Sorted by the width of the pair, the rows narrow from top to bottom and the drivers that move the number most sit where the eye lands first. On Superstore, chairs make the widest pair of full price against discounted sales.

At nothing the ridges are small multiples; high up they sit almost on top of each other and a year fits in a pane. They are drawn back to front, so a ridge is always covered by the one below it and never the other way round.

Split it at the centre and the centre line means half the answers lie to the left. Keep it on the right and the centre separates disagreement from everything else. Leave it out and the shares are rebased on the people who took a side. No reading is right for every survey, so the chart says which one it took.

A task with a completion value turns its bar into a track and fills it, so the progress reads at any opacity. The value is taken as nought to one or as nought to a hundred, decided once for the whole column and stated, because deciding row by row would read a single overshoot as a fraction of a percent.

A marker where the goal sits, the bar filling towards it, and the value, the percentage or both at the end. Past target the track stretches so every bar stays on one scale, and the chart says it did.

Thin stems and small heads stay readable where bars turn into a solid block. Sort them, colour them by a dimension, put them across or down the page.

Every rise in one colour, every fall in another, and the legend counts them. On Superstore between 2022 and 2025, fifteen sub-categories rose and exactly two fell, and the two red lines are the whole story.

Sorting by change buries the biggest collapse at the bottom of the chart. Sorting by the size of the move, in either direction, brings the real movers to the top. On Superstore that puts Copiers first at plus 422 percent and Machines high up at minus 30.

Sigma is estimated from the moving range of consecutive points, so a shift in the process cannot widen the limits meant to catch it. The standard deviation is there when you want it, and the limits carry their numbers.

Cohorts start together on the left whatever month they began in. Periods a cohort has not lived through yet stay empty rather than being filled with a zero, so the triangle is the truth about your data.

Every column is as wide as its share of the grand total. On Superstore, West is 32 percent of sales and South is 17, and the columns are exactly that wide before a single label is read.

A vertical marker sits exactly where the cumulative curve crosses your threshold, and the two colours change at the same bar. On Superstore, nine of seventeen sub-categories carry 83 percent of sales, and the legend says so in real numbers.

Every node splits by the dimension you choose at that node, so two branches of the same tree can follow different paths. Bars and shares keep every level comparable.

Up to four levels of dimensions nest inside each other, every circle sized by its share. The big players are visibly big, and the long tail is visibly a tail.

Each violin carries its density curve, a box with the quartiles and the median, and optionally every observation as a jittered dot. Three views of the same data, one mark.

Drop open, high, low and close on the OHLC shelf, or a single measure: the chart then builds each candle from the first, highest, lowest and last value per day, week or month, choosing the period that keeps the candles readable.

Year, Month and Week pills sit in the toolbar, a month label in the year view opens that month, a week number opens that week. The View setting is only the page the calendar opens on.

The shape is what a reader compares, the number is what they quote, so a column shows both by default. Up to 25 dimensions and 40 measures on one table.

Steps stay in the order Tableau delivers them, because a waterfall is a narrative and sorting by size would tell a different one. Opening and closing bars are one click away.

Drop the dimensions on Levels in the order the flow runs. Two levels give a plain source to target diagram, five give a full path through your data.

Each band carries its stage and its value, and the conversion from one stage to the next is one placeholder away in the tooltip and in the labels.

Sales by hour, incidents by weekday, wind by direction. The first and the last category sit next to each other, which is where the pattern usually is.

Drop several measures for one series each, or a single measure with a dimension for one shape per value. Both shapes of data draw the same chart.

Drag measures to X, Y, and Z, and the axis box renders and scales on its own, with round-value ticks and smart K, M, and B units on every axis. Add a fourth measure to Size, and a dimension on the Detail shelf to color the points by category. With only two measures in play, it renders a classic 2D scatterplot automatically.