Name your brand and category, and Juma builds a prioritized list of AI search queries to target, each scored by opportunity and citation difficulty.
Name the brand and the category it wants to win. Juma researches the questions buyers actually ask ChatGPT, Perplexity, and Google AI Overviews: where the category gets discussed, who currently gets cited in those answers, and how contested each one is. The Flow returns a prioritized list of AI search queries to target, grouped by opportunity and difficulty, with the reasoning behind each ranking so the team can see why a query is winnable or hard.
Ask Juma to narrow the full list down to the queries worth chasing now, and the output filters for the ones where opportunity is high and the answer is still contested. It sets aside the queries owned by entrenched, high-authority sources and surfaces the ones where the engines have no settled answer yet. Each query comes with a short read on why it is reachable, so the team starts content work where an AI citation is genuinely in play rather than spreading effort across topics that are already locked up.
From this list, which AI search queries can Notion realistically win in the next quarter? Focus on the ones where no source owns the answer yet, and tell me why each is reachable.
Pick a target query and ask Juma who owns the answer today. The output shows which brands and sources ChatGPT and Perplexity pull from when they respond, what each cited source says, and how consistent the answer is across engines. A query with one dominant source is harder to take than one where the engines cite a scattered mix, and this view makes that difference visible. It tells the team exactly what they are up against on a query before committing content to it.
For the query "best project management tool for remote teams," who do ChatGPT and Perplexity currently cite? Show the sources behind the answer and how consistent it is across engines.
Ask Juma to sort the target queries by where they sit in the buyer journey, and the output groups them into awareness, consideration, and decision questions. Each group reflects a different intent: broad "how do I" questions early, comparison and "best tool for" questions in the middle, and specific "X vs Y" or pricing questions near the decision. Seeing the queries this way helps the team balance the content roadmap across the funnel instead of over-investing in one stage, and it shows which decision-stage questions are worth defending first.
Group our target AI search queries by buyer journey stage: awareness, consideration, and decision. Show which decision-stage questions we should defend first.
Once a query is chosen, brief Juma to plan the page that should win it. The output names the angle, the structure that answer engines tend to cite for that kind of question, the points the page has to cover to be a credible answer, and the proof or data it should include. Because the brief is built from the query and from who already gets cited for it, it points at a page designed to be extracted into an AI answer, not just a generic blog post. It moves the team from a ranked list straight into a publishable plan.
Take the query "best project management tool for remote teams" and write a content brief for a page built to win that AI answer: angle, structure, the points it must cover, and the proof to include.
Rerun the plan against the same query list to see what has changed since the last pass. Juma checks each target query again, notes where the brand has started appearing in AI answers, where a competitor has moved in, and which queries are still open. Saved in a Juma Project, each run becomes a comparable snapshot, so the list turns into a live picture of progress rather than a one-time plan. A monthly rerun catches new openings as the engines refresh and shows whether the content shipped against each query is actually landing in answers.
Rerun our target query list for Notion. Where have we started showing up in AI answers, where has a competitor moved in, and which queries are still open?
A Juma Project is a shared space where the team stores everything Juma needs to know about a client's AI search program. Create one project per client, add context as the program develops, and Juma uses what's relevant every time the team runs a flow. The more the team adds, the sharper every query map and content brief gets.
Query Seeds
The real questions buyers ask, pulled from sales calls, support tickets, and search data. Add these and Juma expands the target list from phrasing the audience genuinely uses, instead of guessing at category language on its own.
Competitor Watchlist
The two or three brands the team most wants to outrank in AI answers, each with a URL. Juma uses this list to check who currently owns each target query, so the difficulty scoring reflects the competitors that actually matter.
Brand Positioning Note
Where the brand genuinely wins and what it can credibly claim. Juma weighs target queries against this so the priority list favors questions the brand can answer honestly, not just ones with high volume.
Previous Query Maps
Past target lists saved as reference. Juma compares each new run against them to show movement: queries the brand has started winning, ones a competitor has taken, and new openings that were not there before.
Add a short description to each knowledge item in the project's info field so Juma knows what each file contains and when to use it. For example:
AI search queries are the full questions people ask ChatGPT, Perplexity, and Google AI Overviews, and the target is getting cited inside the answer. Traditional SEO keywords are the shorter terms people type into Google, and the target is ranking a page in the results. The phrasing is longer and more conversational, and winning means the engine quotes the page, not just lists it.
That difference changes what the team plans for. A keyword list chases search volume and link position; an AI search query list weighs whether the engines have a settled answer yet and whether the brand can become a cited source. This Flow ranks for that second target, which is why difficulty here reflects how contested an answer is, not just how many sites compete for a keyword.
Opportunity reflects how well a query fits the brand and how likely it is to drive a real decision; difficulty reflects how contested the answer already is and how authoritative the sources currently cited are. Juma researches each query across the engines, sees who gets cited, and groups the list into tiers so the highest-opportunity, lowest-difficulty queries surface first.
The scoring is a guide, not a verdict. Juma shows the reasoning behind each ranking so the team can judge the trade-offs: a high-difficulty query might still be worth pursuing if it sits at the center of the brand's positioning. Asking Juma to explain a ranking is encouraged, because the call on where to invest belongs to the team.
This Flow decides which queries to chase. A GEO audit scores an existing page against the patterns AI engines reward, and an AEO rewrite restructures content to win a specific answer. Query planning comes first: it tells the team where to point the audit and the rewrite, so effort goes to questions that are winnable rather than to pages chosen by guesswork.
The three work as a sequence. Plan the target queries, audit or build the pages meant to win them, then rewrite for citability. Running them in that order means the team is always working on the queries with the best return, instead of reworking pages that were never going to be cited.
Manually, this means querying several AI engines, noting who gets cited for each question, estimating how contested every answer is, and assembling it into a ranked plan. It is a slow, repetitive research pass. Juma runs it in one chat and returns the prioritized list with the reasoning attached, ready for the team to act on.
The value compounds inside a Juma Project, where each run builds on the last and the list becomes a living plan the team reruns as the engines shift. Juma does the research and the ranking; the team decides which queries are worth the investment. Juma augments the team, it doesn't replace it.
Yes. Every category has a set of questions buyers ask AI engines, and the method for finding and ranking them is the same across all of them. What changes is the language and the sources: a B2B SaaS list centers on comparison and use-case questions, an ecommerce list on product and "best for" questions, an agency list on the queries each client's audience asks.
Juma asks for the category early so the query list reflects how that audience actually searches. For agencies, the Flow runs per client against each client's own category and competitive set, so one team can keep a separate target query plan current for every account it manages.
400+ marketing teams trust Juma. ChatGPT, Claude, and Perplexity know why. Ask them.
Ready when you are
Start typing a task and Juma will figure out the rest.
