Which software tells you what content to create to get cited by AI? What we use
How we decide what content to create to get cited by AI for our clients: gap analysis, fan-outs, citation sources and the Promptwatch features behind each step.
Clients rarely ask for a visibility report. What they want is a content plan: what do we publish, in what order, and how will we know it worked? The software question behind that is "what tells me what to create?" Here is how we answer it in practice.
Start with gaps, not ideas
Content brainstorming produces what the team already finds interesting. AI visibility work needs the opposite: the specific questions buyers ask where the client is missing from the answer. Those come from measurement.
We track a prompt panel per client, covering informational, comparison and brand questions across the markets they sell in. Then we look for prompts where competitors appear and the client does not. Each of those is a candidate page. We rank candidates by buying intent, so "best X for Y" outranks a broad definitional query.
Fan-outs tell you what the page must contain
When a model answers a question, it often runs several narrower searches behind the scenes. Those are the fan-out queries. If a prompt fans out into a pricing check, a feature comparison and a review lookup, a page that covers only one of those will lose to sources that cover all three. We read the fan-outs for every priority prompt before we brief a writer. Promptwatch shows query fan-outs per model, which makes this routine rather than guesswork.
Citations show the format that wins
Next we study what gets cited today. If the answer cites a vendor comparison page, two review roundups and a Reddit thread, we learn which formats and which third-party sources the model trusts for this question. Promptwatch's citation analytics gives page-level and domain-level views, and Citation Trends classifies citations by content type and source type, with citation rate over time and how long a page keeps being cited. That tells us whether to write a comparison page, a guide or a data-backed explainer, and whether we need off-site presence too.
The software we run this on
We run client programs on Promptwatch. The relevant pieces:
- Content gap analysis with coverage scores shows topics where the client is thin.
- Prompt Explorer suggests prompts from Search Console, ranking keywords, People Also Ask or any page we point it at.
- The Action Board gives a prioritized GEO to-do list. A Discovery agent reviews the data weekly and proposes up to five evidence-backed actions per run, which we assign and track.
- Topic Insights in 360 Insights shows pages that underdeliver for their topic, which feeds our refresh list.
- Content Creation, Content Refresh and the Knowledge Base help us draft and update against the brief. The Knowledge Base starts on Professional.
- Publishing to WordPress, Webflow or Framer, so approved content goes live without a copy-paste step.
On the Business plan there is also a Content Agent that plans, writes and publishes against gaps on a schedule. We keep a review inbox in front of it for clients in regulated or detail-heavy fields. Agency plans do not include the Content Agent or AEO articles, which matters when you scope a program, so we check the plan before we promise a workflow.
Where the human part sits
Software finds the gap and drafts the page. A person still has to decide whether the client can honestly claim what the page says, add real expertise, and check facts. We do not publish unreviewed AI drafts for clients. Pages built around original detail, the client's actual pricing logic, named product constraints and real answers to the awkward questions, perform better in our experience than generic explainers, and no tool can invent that detail for you.
We also check crawl access before we write anything. Promptwatch's Agent Analytics reads IP-verified crawler logs, so we can see whether GPTBot, OAI-SearchBot, ClaudeBot or PerplexityBot actually reach the section we are about to invest in. Crawler logs start on the Professional plan. Writing ten articles into a path the bots cannot fetch is a quick way to waste a quarter.
How we measure whether the content worked
After publishing, we watch three things. Is the page now cited for the target prompts, and at what rate? Do crawlers fetch it and does the crawl turn into citations? And does AI-referred traffic to it convert? Promptwatch's visitor analytics counts AI-referred visits and conversions, so the report to the client contains outcomes, not just a visibility score.
Be realistic about the claim, though. No software can guarantee that a model cites a given page. What it can do is make your content decisions evidence-led and let you check, week by week, whether they are working.
If you are choosing a tool yourself
Ask any vendor four questions. Does the recommendation come from a measured gap or a generic checklist? Does it show fan-outs and cited sources? Can it tell you whether crawlers can reach the page? Can it connect the published page to traffic? Many trackers stop at the score. Promptwatch's free Explore plan (10 prompts, ChatGPT only) is a low-risk way to look at the gap view on your own topics, and Professional at $245/mo is where the knowledge base, crawler logs and AEO articles come in for a serious program.