Rank Analysis.

A rank question asks respondents to put choices in order — first pick, second pick, and so on. The example here is a hot-sauce U&A study: "Q16. Based on your purchasing decisions for hot sauce, please rank these factors", where 1,003 people ranked their top 5 of ten factors. (New to Analyse? Start with the Analysis Overview.)

Open a rank view and the defaults are:

Rows = Choices, Columns = Ranks — every factor against every position
Values = Cell %

A rank question's default Glow Table: ten purchase factors in rows, Rank 1 to Rank 5 in columns, Cell % values

Two things to read off this table:

Down a column: each rank position splits 100% between the factors — Taste takes 31% of all first places, ahead of type of sauce (15%), heat level (13%) and price (12%).
Across a row: each factor’s Total (n) is how many people put it anywhere in their ranking — itself a finding. 895 of 1,003 ranked Taste somewhere in their top 5; only 155 found room for dietary requirements.

The Target slot — two ways to zoom in

The whole-question view answers "what matters overall". For everything else there’s the Target slot (first above the table) — and for a rank question its menu offers two kinds of target: any choice, or any rank position:

The Target slot menu on a rank question, listing the ranked factors and then Rank 1 through Rank 5

Target a choice — one factor’s full rank profile

Target "Taste" and the view focuses on that factor: its share of each position, with Filters on the other axis so you can add segments:

Q16 targeted on Taste with age and gender cross filter rows: 31% of everyone ranks it first, 35% of 18-35s, 27% of 36-55s

Taste isn’t just ranked often — it’s ranked high: 31% of everyone puts it first — and the segment rows sharpen the story. 18–35s lead the taste-first charge (35%) while 36–55s are noticeably cooler on it (27%).

The age and gender rows are ordinary cross filters: build them once from your demographic questions in Create & Import Filters (here 18–35, 36–55 and 56+ are each an "Any of" over the age brackets), then click the + at the end of the chips to add them — exactly as in Prefilters & Cross Filters. + Prefilter above the chart narrows the whole view to one audience instead — watch n drop in the top corner.

Target a rank — the first-place podium

Target "Rank 1" instead and the table becomes the podium: every factor’s share of first places. Put Choices in the Rows and it reads as a leaderboard:

Q16 targeted on Rank 1 with age and gender columns: every factor's share of first places per segment, led by Taste

309 of the 1,003 first places go to Taste; brand, for all its marketing budget, collects just 102. And the segment columns reshuffle the podium: the heat level jumps to 17% of first places among 56+ (vs 11% for younger buyers), price matters more to men (15%) than women (10%), and packaging design only registers with 18–35s. Cycle the Target through Rank 2, Rank 3… to see how the order changes further down people’s lists, and pick "-" to return to the whole question.

Tidy & sort

On a rank question, both configs travel light: Choices ⚙ and Ranks ⚙ each let you relabel, hide and reorder — ranked items can’t (yet) be merged, excluded or scored, so the full choice toolkit from other question types doesn’t apply here (see Configure Choices, Statements & Loops). The Sort slot still orders the table by results automatically — sort descending by the Rank 1 column and the leaderboard builds itself.

Statistics

With filters in your rows or columns (target mode), tick Subset Significance Test in the dashboard Settings (⚙) to highlight segment differences green or red against the Total at a 95% confidence level, and Margin of Error to size up each base. Details: Significance Testing and Margin of Error.

When the analysis looks right, click Save — switching dashboards discards unsaved changes.

FAQs

Why is every factor’s Total (n) different?
A factor’s base is the people who put it anywhere in their ranking. If respondents rank their top 5 of ten factors, the ones left off someone’s list don’t count for them — so universally-ranked factors (Taste, 895) carry bigger bases than niche ones (dietary requirements, 155).

Why do the columns only go to Rank 5 when there are ten factors?
The question only asked for a top 5 — positions that were never offered can’t appear. The table always mirrors what respondents actually did.

How do I rank the table by first places automatically?
Use the Sort slot above the table — sort descending keyed to the Rank 1 column — rather than reordering by hand in the config.

Where did "Add Global" go?
Retired — and it never added filters to "all views in your survey", only the dashboard. The Add to menu now offers Current View, Selected Views or Global (every view in the dashboard), and chips stay ordinary local chips. See Prefilters & Cross Filters.