Read all four intents side by side
Open one view and see Informational, Commercial, Navigational, and Transactional clicks, impressions and position next to each other. No more guessing which kind of search drives the result.
Read your search demand by intent
Clicks went up. Buying intent might not have moved at all — Search Console tells you what people search, never why.
Search Console tells you what people search, never why. We classify every query with an LLM in your own warehouse: Informational, Commercial, Navigational and Transactional. We then report clicks, impressions and CTR by intent, instead of one row at a time.
Intent ClicksImpr.Avg. pos
Informational
—
—
—
Commercial
—
—
—
Navigational
—
—
—
Transactional
—
—
—
Illustration. Each intent carries its own clicks, impressions, CTR and average position, computed from your own queries.
Open Search Console in another tab. Can you answer these right now?
- Of your top queries, how many are transactional — someone ready to act — versus just researching?
- When traffic grows, is it buying-intent searches climbing, or informational volume?
- That high-impression, low-CTR query — is it a problem, or just people learning, not buying?
If you hesitated on any of these, the answer is already in your data — just not in a shape Search Console will show you.
Built for teams who own their data
- Northpeak
- Datawise
- Quantly
- Meridian
- Brightline
- Corewave
Why you can trust the data
-
Read-only access
We only ever read your data. Nothing on your site or in Search Console is ever changed.
-
Your warehouse
All data lives in your own Google Cloud project. It is never copied into a tool you do not control.
-
Your rules
Grouping and classification are tailored to your site, and stay fully editable as it changes.
-
Yours to keep
Leave and the tables stay. Export on your terms, with no rolling deletion window and no lock-in.
-
Founder pedigree
Built by a team that has run analytics for high-traffic platforms, with a Google Product Expert among the founders.
-
Built on Google’s stack
BigQuery storage and the full Search Console API, not a third-party scrape.
You have said one of these, but could not tell research from revenue
Sort your queries by clicks and the informational ones win the top — they always carry the volume. So the report flatters the wrong searches: high-traffic curiosity looks like success while the smaller commercial and transactional queries that actually turn into revenue stay buried, weighted identically to everything else.
The number looks like growth; the buying intent behind it may not have moved at all.
In-house, measured on traffic
I chase the queries with the most volume, we rank and the clicks come in. Then conversions do not move. The report does not tell me which of these searches are people ready to buy and which are just reading. So I keep pouring effort into terms that look big and do not earn.
The real problem A query is just a phrase with clicks. Nothing marks whether the searcher is learning or buying. So high-volume research outranks the smaller converting queries in every click-sorted view, and effort follows volume instead of intent.
Editorial / content marketing
The query report is mostly ‘how to’ and ‘what is’, so we keep shipping guides. The ‘best X’ and ‘buy X’ searches are the ones close to a decision, and they barely register. I have no way to see they are under-served. We are feeding the top of the funnel and starving the part that converts.
The real problem Without the intent behind each query, the high-volume informational terms dominate what looks worth writing. So commercial and transactional demand stays an invisible thin slice that no one briefs a page for.
Reporting to clients
I show the client we now rank for two hundred new queries and clicks are up. Then they ask whether any of that actually sold anything. I cannot separate the research traffic that pads the report from the buying traffic that moves their revenue. So the win I am presenting might be all top-of-funnel.
The real problem The report counts clicks without the intent behind them. So there is no way to show whether the growth is converting demand or just research volume. A rise in traffic cannot answer the question the client actually asks.
Exec, owns the budget
The SEO numbers look strong but they do not show up in pipeline, and I cannot tell why. Is search just failing to reach people ready to buy, or is the buying demand there and we are losing it? Without that, I cannot decide whether to put more budget into SEO or pull it.
The real problem With no intent on the queries, the report cannot say whether SEO is reaching commercial and transactional demand at all. So a big traffic number gives the exec no read on whether it connects to revenue, and the budget call is a guess.
How working with us works
A managed setup, scoped to your site. It is not a plan you configure yourself.
- 01
Discovery call
We look at your queries and Search Console together, and agree on the intent rubric that fits your site.
- 02
We build it in your BigQuery
We classify your full query history with an LLM inside your own warehouse, read-only.
- 03
It runs every day
New queries are classified automatically and every report stays current. Nothing for you to maintain.
A managed setup: you give read-only access, we run it, and your data stays in your own warehouse.
Performance, split four ways
Each period your queries collect into four intents. The clicks, impressions and CTR for each tell you which demand is research and which is ready to act.
- Informational 58% of queries
- Clicks
- 4,210
- Impressions
- 182,400
- CTR
- 2.3%
- Commercial 22% of queries
- Clicks
- 1,980
- Impressions
- 38,600
- CTR
- 5.1%
- Navigational 12% of queries
- Clicks
- 3,140
- Impressions
- 9,800
- CTR
- 32.0%
- Transactional 8% of queries
- Clicks
- 1,260
- Impressions
- 14,200
- CTR
- 8.9%
Example rollup. An LLM classifies your intents from your own Search Console queries.
What you actually do with it
In the order you’ll use it, from the first read to the report your client opens.
Watch an intent’s share move, not a single query
See a whole intent’s line shift over time. A slide toward research, or away from buying demand, shows up while you can still act on it.
New queries classified automatically
A query you’ve never ranked for is labelled the first time it appears in Search Console. No re-tagging.
Labels you can see and tune
The LLM’s labels live in your own BigQuery next to the raw metrics. Review them, adjust the edge cases and trust what rolls up. It is a system you can inspect, not one hidden in someone else’s dashboard.
In every report, not one dashboard
Because the intent lives in your data layer, it joins Looker Studio, your BI and the digest email. It reaches everywhere your numbers already go.
The difference from doing it by hand
You can tag intent yourself in a spreadsheet, until the row cap, the hours and the inconsistency catch up. Here is what changes when an LLM classifies it in your warehouse instead.
By hand in a spreadsheet
Manual, session-bound
- Tagging every query
- By hand Hand-label ~1,000 rows, 2 to 4 hours
- Insightlytics An LLM labels every query
- Coverage
- By hand The top rows you can export
- Insightlytics Every query, no row cap
- History
- By hand Only the last 16 months
- Insightlytics No limit, kept as long as you need
- Consistency
- By hand Subjective, drifts per person
- Insightlytics One model, one rubric
- Ambiguous phrasing
- By hand Keyword rules guess wrong
- Insightlytics Read in context by the LLM
- Performance by intent
- By hand Re-pivot the export by hand
- Insightlytics Joined into every report
- Keeping it current
- By hand A manual pass each quarter
- Insightlytics Re-classified automatically
Pricing
Priced to your site, not a plan you pick
No tiers to choose between. Pricing depends on your site's size and data. We scope it with you on a quick call, and your data is always yours.
Frequently asked questions
If something is still unclear, a discovery call clears it up fast.
Informational (learning something: ‘how to’, ‘what is’, guides), Commercial (comparing before deciding: ‘best’, ‘review’, ‘X vs Y’), Navigational (heading for a specific site or brand by name) and Transactional (ready to act: ‘buy’, ‘price’, ‘free trial’, ‘download’). Every query is sorted into one of the four.