The problem this solves
The keyword gap - terms your competitors rank for while you are absent - is the most concrete map of forfeited demand a content team can hold. Producing it properly is fiddly: export rankings for three domains, dedupe tens of thousands of rows, join volumes and difficulty, filter out branded noise, then decide what is actually winnable. Most teams do it once during an agency audit, act on a fraction of it, and never refresh it.
The result is a content calendar steered by instinct while competitors quietly own the queries with commercial intent. A gap analysis that arrives as a raw export has a second problem: ten thousand rows is where prioritization goes to die. What planning actually needs is the full dataset for reference and a short, argued list of the plays worth running this quarter - both at once, from the same run, so the list can always be traced back to the data.
How the mission runs
- Pull three ranking footprints. Keyword Intelligence retrieves the ranked keywords for both competitor domains and your own, with position, search volume, and difficulty per term. Running all three in one mission matters, because the gap only exists relative to where your own domain already stands.
- Compute the true gap. The agent joins the datasets and isolates terms where either rival ranks and your domain is absent or buried. Branded terms, junk queries, and irrelevant geographies are filtered out, so the gap reflects demand you could plausibly contest rather than raw overlap noise.
- Score by winnable value. Each gap term is scored on volume, difficulty, commercial intent, and how strong the ranking competitor's page actually is. The ranking favors terms where effort and payoff meet - meaningful volume, beatable pages, and intent that maps to revenue. The scoring logic is written into the output so you can audit or reweight it.
- Ship the CSV and the brief. The complete gap lands as a CSV with every column needed to slice it your own way. On top sits a written brief arguing the ten best plays: the term cluster, why it is winnable, what the page should be, and which rival you would displace.
The prompt
This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.
What comes back
Two artifacts from one run: a CSV containing the full keyword gap - term, volumes, difficulty, each rival's position, intent tag, and winnability score - and a short brief that walks through the top ten opportunities with the reasoning spelled out. The CSV feeds whatever planning sheet you already use; the brief is ready to paste into a content planning doc or a sprint ticket.
Make it yours
- Restrict the gap to bottom-funnel modifiers such as pricing, alternatives, and comparison terms when the goal is pipeline influence rather than raw traffic volume.
- Run it against five rivals and ask for terms where at least two of them rank, which surfaces the queries your whole category has decided are valuable.
- Re-run quarterly and have the brief open with gap terms closed since last time, turning the mission into a scoreboard for the content program.
Frequently asked questions
How fresh is the ranking data?
Keyword Intelligence reflects the most recent crawl of each domain's rankings at the time the mission runs. Rankings move constantly, so the CSV stamps its retrieval date, and the practical answer is to re-run the mission each quarter so planning always works from a current gap.
How does the agent decide what is winnable?
It weighs volume against difficulty, reads the intent of each term, and inspects the strength of the pages currently ranking. A high-volume term behind entrenched pages scores lower than a moderate term with weak coverage. The scoring rationale ships in the brief so you can challenge it.
Can I review the brief before it circulates?
The CSV and brief land as mission artifacts for you first, and nothing is sent anywhere until you move it. Most teams review the top-ten list, adjust a pick or two with context the data cannot see, and then share it in their planning channel.