Skip to content
SilktideHelp

Heatmap

The heatmap lives on the Presence Overview, between the headline cards and the standings table. The screens answer one question at a time: who wins this topic. The heatmap widens that across everything you measure, so you can see the shape of the whole competitive picture at once.

It answers two questions the single-arena screens cannot:

  1. Where am I weak? Which arenas is your brand losing that you assumed you were winning?
  2. Who is quietly taking everything? Which competitor is strong across arenas rather than in the one place you already watch?

Brands run across the top and arenas run down the side. There are usually only a handful of brands and a brand compresses to a logo, whereas topic names are sentences that need room to be read.

The Brands filter in the toolbar chooses who is in the comparison. All of them by default; narrow it to two or three when you want a head-to-head without the rest of the field in the way. Everything below re-scales to what is left, including the shading and the Overall row.

The two measures

TabWhat each cell means
MentionsThe share of that row's measured answers that named the brand.
CitationsThe share of that row's cited pages that belong to the brand.

These are deliberately separate rather than blended into one score, because they disagree, and the disagreement is the finding. Being named constantly while nobody cites your pages is a different problem, with a different fix, from being cited by everyone and named by nobody.

Citations need a web address on each brand before they can be attributed. If no tracked brand has one, the Citations tab explains that instead of showing a grid of zeroes. Adding addresses fills in the whole history, with nothing to re-run.

Rows: the workspace's topics

The rows are the of the you have selected in the toolbar. The whole screen is one workspace, so there is nothing coarser to group the rows by - grouping them back up would collapse the grid to a single line. Narrowing the filter to a hand-picked set of topics narrows the rows to match.

Click a cell to open the Responses list for that topic and brand. Click a topic name to open every response in that topic. Citations cells are the exception: they open the cited pages that produced the number.

Reading the shading

Shading is relative to the strongest cell currently on screen, not to an absolute 0-100%. Real answer sets cluster low: a 30% share is often the runaway leader in its arena, and an absolute scale would render the whole grid the same pale wash. This does mean the colors re-scale when you change the filters, so the legend always states the range it is currently using.

The number is printed in every cell, so the color is a scanning aid and never the only signal. Shading fades in from transparent to solid along a three-stop hue walk (mentions: blue through violet to red; citations: cyan to purple) - never a red-to-green scale - so the order still holds for readers with any form of color blindness.

Three cell states are deliberately distinct:

CellMeaning
A shaded percentageThe brand appeared in that share of the row's answers.
A plain 0% on a flat backgroundThe row was measured and the brand never appeared.
A hatched cell reading n/aNothing was measured for that row in this window.

The difference between the last two matters. A measured zero is a result you can act on; an unmeasured cell just means you have no running there yet.

Sorting

A grid has two axes, and both of them sort, independently of each other.

Sorting by a column reorders the arenas. Every heading sorts: the arena name, the answer count, and each brand. Clicking a brand sorts the arenas by how that brand does in them, which is the fastest way to answer "where is this competitor beating me". The grid opens sorted by your own brand, strongest arena first.

Sorting by a row reorders the brands. The arrow beside each row name sorts the brand columns by that row's figures. Clicking it on "PDF" answers "who owns this arena" without reading across the line. The grid opens sorted by the Overall row, strongest brand first.

The arrows point along the axis they move things, and in both cases they point the way the figures fall. A column sorted descending shows a down arrow, because the numbers get smaller as you read down it. A row sorted descending shows a right arrow, because the numbers get smaller as you read across it, which puts the leader on the left.

Your own brand is not forced to the front. You choose it in the toolbar and the grid opens ordered against it, but it takes its place in the ranking like everyone else, because a comparison you have quietly rigged in your own favor is not a comparison.

Nulls always sort last, in both directions, on both axes: "no data" is not a low score. Ties break alphabetically, so the order is stable rather than arbitrary.

What is excluded

are left out of the heatmap entirely. Naming yourself in an answer to "who is Silktide?" is not a competitive win, and including those topics would inflate every owned brand's row.

The Overall row at the foot is each brand's figure across everything currently in the grid, so it moves with your filters.

Last updated

Was this page helpful?

Heatmap | Silktide Help