Find Keyword Opportunities
Keyword research answers one question: which searches does this site have a realistic shot at? Not the biggest searches, and not the easiest - the ones where genuine demand, winnable competition, real relevance to your business, and a page worth linking to all overlap.
Before you start
Research is grounded in what your site actually covers, so it needs something to read:
- At least one atomized page on the domain. More is better - it is what the scoring uses to judge relevance, and what the target matching picks between.
- Optionally, Google Search Console connected for that domain. Not required, but it adds the queries you already appear for, which are the strongest candidates on the list.
Run it
- Go to Opportunities and pick the domain.
- Choose how many keywords to look for (100 to 500).
- Click Run research.
It takes a few minutes. When it finishes you get a headline - "We found 137 search opportunities for your site" - and the list below it.
Re-running is safe and additive. Metrics refresh in place, new demand gets added, and anything you have dismissed or already targeted keeps its state. It is not a reset.
Reading the list
Opportunity bands sort everything into three piles:
| Band | Means |
|---|---|
| High | Worth doing now - strong relevance, winnable difficulty, and a page that can carry it |
| Medium | Worth a try - real demand, but one factor is working against it |
| Exploratory | Long shots - thin demand, weak relevance, or a results page you are unlikely to break into |
Two separate lists. Long-tail opportunities appear first, head and category terms below them. That split is deliberate: the score rates how winnable a query is, and head terms score lower because broad results pages are genuinely harder to break into. That does not make them less valuable - on a site named after its category they are often the most valuable demand you have. One combined list sorted by score would bury them under hundreds of longtails that scored better by being easier.
Click any keyword to see exactly how its score was reached - every factor, its weight, and a sentence explaining it. Nothing is a black box.
The seven scoring factors
| Factor | What it measures |
|---|---|
| Search volume | How many people search it each month |
| Difficulty | How contested the results page is |
| Relevance | How squarely the query is about your business |
| Commercial intent | Whether the searcher is close to buying |
| Existing coverage | Where you already rank - position 11-25 is striking distance, the most actionable state there is |
| SERP fit | Whether a third-party placement can rank for this shape of query at all |
| Longtail attainability | Whether the phrase is winnable without years of accumulated authority |
Relevance is a gate, not just a weight: a query your business has no real standing in never reaches a recommended band, however large it is.
Where the numbers come from
Every figure is labelled with its source, and the label matters:
1,400/mo- measured by a connected keyword data provider.1,400/mo on Bing- measured by Bing Webmaster Tools. Bing is a fraction of Google's search share, so treat it as a floor on real demand, not a total.~90/mo- estimated from your own Search Console impressions. Directional, not measured.—- nobody measured it. The score still works; it leans on relevance, query shape, your existing coverage, and whether search autocomplete confirms people type the phrase.
Nothing here is ever invented. If a number is not shown, it is because no one measured it.
"Page too thin"
Some opportunities show a warning instead of a target page:
Search opportunity identified, but your page currently lacks enough topical coverage to compete.
This is a finding, not an error. The demand is real; the page you would send people to cannot answer the query. Publishing anyway would produce posts promising an answer and a link that does not deliver it.
Two ways forward:
- Improve the page so it genuinely covers the query, then re-run research. The opportunity releases itself.
- Write the page the query deserves, then atomize it.
Triage
- Target this builds a keyword campaign. → Build a Keyword Campaign
- Dismiss removes it for good - later research runs never resurface it.
Hand it over entirely
Under Find and target automatically:
- Auto-target keywords - each month, build campaigns for the best opportunities without asking. Only keywords whose target page can already carry them are picked up, so nothing gets published at a page that would lose.
- Keywords per month - how many new campaigns to start. Post credits still bound what actually gets generated.
- Minimum opportunity score - the bar a keyword must clear. 70 and above is the High band.
- Refresh research monthly - re-run discovery about every four weeks so new demand shows up on its own.
Availability
Opportunity Mode is included on Growth and above. → Plans and Pricing