Ask the questions Search Console cannot answer.
Sixteen reports over the full history of your site — queries, pages, markets and SERP features, answered inside Prism.
- backfilled on connect, then kept for good
- 16mo
- backfilled on connect, then kept for good
- to connect — the backfill starts on its own
- 2min
- to connect — the backfill starts on its own
- reports over the same warehouse
- 16
- reports over the same warehouse
- credits used by any of them
- 0
- credits used by any of them
Every row opens into its own history.
Ask how a query has moved over two years and the answer is already there — each history is read from your own stored record, not from whatever window Google is currently serving.
- Keywords
- Every query in one table; a row expands into its stored position history.
- Position Changes
- Risers and fallers together, each against its previous-period average.
- Cannibalization
- Your own pages competing for one query, matched on meaning rather than spelling.
Opportunities reads the data and hands you the work.
The shortest route from the numbers to next week's plan — three lists, rebuilt as each day's data lands.
- Striking distance
Queries sitting just off page one, where one improvement moves real traffic.
- Below the curve
Pages earning less than your own click-through curve says they should.
- Sustained declines
Slides that took six months, caught while they are still cheap to reverse.
Sixteen reports, four jobs.
The whole suite, grouped by what you open each report to find out. All sixteen read the same stored record.
The record
What your search traffic is, read straight from the stored data.
- Overview
- Clicks, impressions, CTR and position for the period, a trend line against the one before, and a panel naming the queries that drove the change.
- Keywords
- Every query, in one table. A row expands into that query’s stored position history, or opens onto a page of its own carrying every URL that has ranked for it and the dated page swaps between them.
- Top Pages
- The pages earning your traffic, each with its leading query, plus traffic by site section and the biggest page movers.
- Devices & Markets
- The same traffic split by country and by device, day by day, with CTR set against position for every market.
Movement
What changed against the previous period, and on which day.
- Position Changes
- Queries whose Google position moved, risers and fallers together, each against its previous-period average.
- Position Distribution
- How many queries sit in each ranking bucket, week by week, and the clicks each bucket brings in.
- Click Gap
- The queries you gained and lost this period, and what each one was worth in clicks.
- Calendar
- Every day plotted, with the weekday rhythm beside it and any day far off its trailing norm flagged as a spike or a drop.
Opportunity and early warning
The work worth doing next, and slippage caught while it is small.
- Opportunities
- Queries in striking distance of page one, pages earning less than your own CTR curve says they should, and pages in sustained decline.
- Cannibalization
- Queries where several of your pages split the impressions, with each contender’s share of them.
- Query Counting
- How many queries each page gets served for. Demand often falls here before clicks do.
- Index & Serving
- Impression shifts by search type, which show indexing and serving changes before rankings do.
How you are found
The shape of the demand behind the numbers.
- Query Intent
- Queries classified as informational, commercial, transactional or navigational, and how each kind performs.
- Branded Search
- People searching your name against people discovering you, split using your brand terms and aliases.
- Keyword Clusters
- Queries grouped by what they mean, or by your own saved topics, with clicks and position per cluster.
- SERP Features
- The special result types your pages appear in and the click share each takes, fetched live from Search Console.
Every page that has ranked for it, and the day it handed over.
Open any row in Keywords and that query gets a page of its own — its full stored history, every URL Search Console has returned for it, and how it performs by market, by device and against your own click-through curve.
- Page swaps, dated
- The day one of your URLs took the query over from another, with each page’s share of the impressions either side of the change.
- Confirmed, not guessed
- A change is reported once the incoming page has led on 7 of the next 10 days that carry data. A one-day flip is left out, and a change too recent to judge is marked pending.
- The pages ranking now
- Status code, indexability and Lighthouse score for each URL currently returned, read from your last crawl.
Google stays the source. Prism is the memory.
Every position here is the average Search Console reports for your own queries, kept permanently — Google’s own record, not a scrape of a results page.
Your first 16 months are everything Google still holds on the day you connect, and every day after that gets added on top.
Frequently asked questions
Which report do I open first?
Opportunities. It puts the queries sitting in striking distance of the first page next to the pages performing below their own click-through curve. Click Gap next, for what moved and where it went, then Cannibalization for the pages competing with each other.
Which reports are included?
Sixteen. Overview, Opportunities, Calendar and Click Gap; Keywords, Position Changes, Position Distribution, Keyword Clusters, Cannibalization, Query Counting, Query Intent and Branded Search; Top Pages and Devices & Markets; SERP Features and Index & Serving. All of them come with the free plan, and none of them uses credits.
Can I look at a single query on its own?
Yes. Every row in Keywords has an Analysis button, and it opens that one query on a page of its own: its stored history over 28 days to 16 months, every URL Search Console has returned for it with the first and last date each appeared, the dated moments the ranking page handed over, the split by country and device, its click-through rate against your own curve at that position, the queries nearest to it in meaning, and the crawl state of the pages ranking for it today.
How do you decide the ranking page has actually changed?
The incoming page has to be the highest-impression page for the query on at least 7 of the next 10 days that carry data, counted in days with data rather than calendar days. Below that the change is treated as noise and nothing is reported, and a change inside the last few stored days is marked as pending instead of being counted either way. Prism dates the change and names the two URLs; it does not attribute a cause, because nothing in the stored data separates a canonical change from a content edit from a Google-side re-pick.
Can I scope a report to part of the site?
Yes. A segment is a saved query group or URL group, and once you pick one in the toolbar every Search Performance report stays scoped to it. Brand against non-brand is the one most people build first.
How is this different from Search Console itself?
Search Console shows you 16 months of data through one interface. Prism keeps that data permanently in its own database and puts 16 reports over it, so questions Search Console cannot answer, like which queries have been sliding for six months, become a page rather than a spreadsheet.
Where do the positions come from?
Straight from Google. They are the average positions Search Console reports for your own queries, kept permanently, so your position history is Google's own record rather than a scrape of a results page. Search Console stays the source; Prism is the memory around it.
The rest of the platform
- AI Visibility See exactly what ChatGPT says about your market: your prompts put to ChatGPT, Claude, Gemini and Perplexity on a schedule, every answer kept word for word, and the brands named on the days you were not. 18 reports.
- Site Health Severity says how alarming a crawl finding looks. Fix First classifies the same findings by what each one is doing to your traffic: broken while earning, recoverable losses, upside, housekeeping. 13 reports over one crawl.
The best day to start keeping your data was 16 months ago.
Connect a property and Prism pulls back everything Google still holds, then keeps every day that follows.