Key takeaways
- Between Sep 28 and Oct 4, 2026, product pages (22.4%) and landing pages (16.8%) together earned about 39% of ChatGPT citations. Most editorial calendars are built almost entirely around blog posts.
- Listicles, how-tos and comparisons add roughly 28%. They matter, but they are the second layer, not the foundation.
- Content type depends on intent. Commercial prompts lean on product pages, landing pages, listicles and comparisons (almost 70% combined). Branded prompts lean on landing pages and reviews.
- Shelf life differs by format. Landing pages and docs stay cited longest. News loses half its pages within 36 days.
- The practical move: split your calendar into evergreen assets you maintain, editorial pieces you refresh, and short-burst content you accept will decay.
Why content type matters more than topic calendars suggest
Most editorial calendars are organized by topic and keyword. Monday is a blog post about X, Thursday is a roundup about Y. Format is an afterthought, usually "blog post" by default.
ChatGPT does not treat formats equally. Promptwatch classifies every citation by content type, and Promptwatch Data's content type report for Sep 28 to Oct 4, 2026 shows how lopsided the mix is. The top two slots belong to pages that a typical content team does not put on the calendar at all.
I find this a little uncomfortable. A lot of content teams, including ones I would call good, spend 80% of their hours on the formats that earn about 43% of citations, and treat product and landing pages as something the web team handles once a year.
The numbers: what ChatGPT cites most
Across all prompts in the Sep 28 to Oct 4, 2026 window, the breakdown looked like this:
| Content type | Share of ChatGPT citations |
|---|---|
| Product pages | 22.4% |
| Landing pages | 16.8% |
| Listicles | 13.6% |
| How-tos | 8.7% |
| Documentation | 7.5% |
| General articles | 6.1% |
| Comparisons | 5.9% |
| News | 5.7% |
| FAQ | 3.3% |
| Reviews | 1.7% |
Two things stand out. Product and landing pages together are about 39%. And all the editorial formats (listicles, how-tos, articles, comparisons, news, reviews and others) add up to about 43%. Listicles, how-tos and comparisons alone earn about 28%.
So the answer to "should we keep publishing editorial content?" is yes. But the answer to "is that the whole job?" is no.
The mix changes with intent
The all-prompts average hides the most useful part of the report. The mix shifts a lot depending on what the user is asking.
Commercial prompts
For prompts like "best CRM for small businesses", product pages take 26.0%, landing pages 20.6%, listicles 15.0% and comparisons 7.9%. Those four types make up almost 70% of commercial citations.
If your calendar targets bottom-funnel queries with how-to blog posts, you are showing up in the wrong format.
Informational prompts
For prompts like "how does a CRM work", the spread is wider. Product pages are still first at 19.1%, then landing pages (13.9%), listicles (13.6%), how-tos (10.7%), documentation (8.5%) and articles (7.7%). This is where your educational content actually competes, and even here it shares the stage with commercial pages.
Branded prompts
For prompts like "is HubSpot good for startups", landing pages jump to 22.8% next to product pages at 23.3%. Reviews reach 5.6%, more than three times their overall share. FAQ pages hit 4.8%, social posts 3.0% and press releases 2.1%.
This matters for any brand question. When someone asks ChatGPT about you by name, the answer is built largely from your own pages and from third-party reviews. A clear FAQ page and a healthy review profile are editorial work too.
| Prompt type | Top formats | What it means for the calendar |
|---|---|---|
| Commercial | Product, landing, listicle, comparison (~70%) | Fix commercial pages first, then get into roundups |
| Informational | Product, landing, listicle, how-to, docs | Publish step-by-step guides and keep docs public |
| Branded | Product, landing, reviews, FAQ | Collect reviews, build a brand FAQ |
Shelf life: which formats stay cited
Citations are not forever. Promptwatch Data's content half-life study took every page ChatGPT cited on July 6, 2026 and tracked it through Oct 4, 2026.
- Landing pages: 78% still cited after 30 days, 69% after 60.
- Docs: 76% after 30 days, 67% after 60. ChatGPT cited 93% of docs again after day 0, against 77% of news pages.
- Comparisons: 72% after 30 days, 63% after 60.
- Listicles: 60% after 60 days.
- Product pages: 55% after 60 days. Product pages, reviews and how-tos all reach half-life at 74 to 77 days.
- News: almost a quarter of pages are never cited again after day 0, half are gone after 36 days, and 41% are still cited at day 60.
Landing pages, docs, comparisons and listicles only crossed the 50% line in the final week of the window, so their true half-life is likely close to or beyond 90 days. Note that the lines drop faster at the end partly because the window closes.
Pricing pages are classified as landing or product pages, which makes them long-term assets. Most teams ignore them after launch.
The trend is not static
The mix moves. Promptwatch Data's August 2026 citation types report showed product pages averaging about 29% of daily citations, holding near 30% until Aug 22, then falling to about 25%. Landing pages fell from about 20% in week one to under 12% at month end. Meanwhile how-tos rose from 4.3% to 9.1% over the last nine days, documentation went from 3.3% to 8.2%, and listicles ended the month at 11%.
Social posts dropped from 4.4% to under 1% after August 14, the same day Reddit's share of ChatGPT citations collapsed from roughly 4% to 0.5%. If you were counting on forum and social presence to get cited in ChatGPT, that bet got worse overnight. The social citations report puts ChatGPT at 0.46% of citations from social platforms for Sep 28 to Oct 4, against 12.2% for Google AI Overviews. Different engines, different calendars.
The lesson is not to chase every weekly swing. It is to avoid putting all your effort in one format, because the balance shifts.
Turning this into an editorial calendar
Here is a way to restructure planning. It is an opinionated starting point, not a law.
Layer 1: evergreen commercial assets (maintain, do not just publish)
Product pages, landing pages, pricing pages and docs. These earn the largest share of citations and last the longest. Put them on the calendar as recurring maintenance tasks:
- Quarterly review of specs, pricing and target audience on every product page.
- Make product pages easy to compare, with clear pricing, specs and who the product is for.
- Keep docs public and crawlable. Gating them removes you from the 7.5% docs slice.
- Add a brand FAQ page if you do not have one.
Layer 2: comparison and roundup content (build and refresh)
Listicles and comparisons are about 19.5% of citations combined, and they stay cited for months. Two actions:
- Publish your own comparison pages, including honest ones against your closest competitors.
- Find the "best of" roundups ChatGPT already cites for your category and work out how to get listed in them. That is outreach and PR time, and it belongs on the calendar.
Layer 3: how-tos and educational guides (steady cadence)
How-tos earn 8.7% overall and 10.7% on informational prompts. They are rising. Publish step-by-step guides with real specifics, and refresh them on a schedule.
Layer 4: news and PR (short bursts, no illusions)
News lost half its cited pages in 36 days. Use it for moments of visibility and do not expect it to compound. Plan it around launches and announcements, and keep the budget proportionate.
A rough rebalancing
| Calendar slot | Typical team today | Suggested direction |
|---|---|---|
| Blog and articles | Most of the hours | Fewer net-new posts, more refreshes |
| Product and landing page upkeep | Ad hoc | Scheduled quarterly reviews |
| Comparisons and own roundups | Occasional | Monthly or bi-monthly |
| Docs and FAQ | Product team only | Shared ownership with content |
| Reviews and third-party listings | Not planned | Standing outreach task |
| News and PR | Reactive | Reserved for launches |
I would not tell anyone to stop writing blog posts. I would tell them to stop treating blog posts as the only unit of content.
Tools for planning and measuring this
You do not need a new stack, but a few categories help.
Tracking what AI actually cites
To see which of your pages get cited and in which format, you need an AI visibility tracker. Promptwatch classifies every citation into content types and tracks first cited date, peak and decay per page, which is the same lens as the half-life data above. It also has content gap analysis and a Content Agent for acting on what it finds. Other trackers worth a look include Otterly.AI and Peec AI, which are more monitoring-focused.
Otterly.AI
For a wider survey, the GEO software directory at bestgeosoftware.com lists more options.
Content briefs and optimization
For the editorial layers, tools like Clearscope, Surfer SEO and Frase help with briefs and on-page optimization.
Landing page production
If landing pages are going on the calendar, a builder like Unbounce or Instapage handles the production side, though the citation value comes from clear content, not from the builder.
Page-change monitoring
To keep tabs on your own and competitors' product and pricing pages, a change detection tool like Visualping is useful.
| Need | Example tools |
|---|---|
| Citation and content type tracking | Promptwatch, Otterly.AI, Peec AI |
| Briefs and optimization | Clearscope, Surfer SEO, Frase |
| Landing page building | Unbounce, Instapage |
| Competitor page monitoring | Visualping |
What the generic calendar advice misses
Much of what ranks for "ChatGPT editorial calendar" is about using ChatGPT to generate a posting schedule. Prompts for 31-day social calendars, content mixes of storytelling and tips, that kind of thing. That is a different question. It asks how to use ChatGPT to plan content. This guide asks what ChatGPT cites once the content exists. A calendar produced by prompting a chatbot will not tell you that product pages out-cite how-tos by more than two to one.
Use those prompt guides for ideation if they help, but ground the format mix in citation data.
A short checklist
- Audit your product, pricing and landing pages. Are specs, prices and audience current?
- List the roundups in your category that ChatGPT cites. Pitch them.
- Publish or update one comparison page this month.
- Check that docs and help content are public and crawlable.
- Add a brand FAQ and a review collection process.
- Move news and PR to a launch-only slot.
- Re-run the numbers monthly. The mix shifted inside August alone.
Caveats worth keeping in mind
These figures describe ChatGPT citations for tracked prompts in specific windows: Sep 28 to Oct 4, 2026 for the mix, and July 6 to Oct 4, 2026 for half-life. A single week is a snapshot. Industries differ too. Promptwatch Data has separate reports for software, travel, healthcare and automotive, where the leading source types differ. For software prompts, documentation and help centers earn the largest share. For automotive, marketplaces outrank every car brand. Check your own vertical before copying a mix.
Also, other engines behave differently. Google AI Overviews cites far more social content than ChatGPT does. A calendar tuned only to ChatGPT will be lopsided elsewhere.
Bottom line
For the week of Sep 28 to Oct 4, 2026, ChatGPT drew about 39% of its citations from product and landing pages, and the pages that last longest are landing pages, docs and comparisons. Put those on your calendar as maintained assets first, add comparisons and roundups second, keep how-tos on a steady cadence, and treat news as a short burst.