Every content team starts the year with a beautiful editorial calendar in Notion or Airtable. By week three of February, half the rows say “TBD — find topic.” By week five, the calendar is abandoned and someone’s asking ChatGPT for “blog post ideas about productivity” on a Sunday night.
The failure is rarely discipline. It’s a supply problem. Topic ideas don’t arrive on a regular schedule, and when you run out, you write filler. Filler is what kills your calendar.
The five-source rule
We spent 18 months experimenting with topic-research workflows across our own sites. The single change that worked: stop relying on one source. Topic ideas have to come from at least five places, refilled continuously, surfaced in one library.
Here are the five sources, in rough order of conversion to actually-published articles:
1. Google Search Console — the queries you almost rank for
Pages 2–3 of the SERP for queries you already rank on. These are signals from the algorithm that you’re close. Improve the post that almost ranks, or write a sharper companion piece. Highest conversion source — these topics already have impression volume.
2. Live SERP — what your competitors are publishing
Pull the top-10 results for your target keyword. Read what they wrote. Find the angle they didn’t cover, the data point they cite without sources, the section every competitor has but none does well. That’s your topic.
3. People Also Ask boxes — search intent in question form
PAA is Google telling you, verbatim, what people are confused about. Each PAA box is 4–8 free article ideas. Most teams ignore them because pulling them manually is tedious — automate it.
4. Reddit — what people complain about, in their own words
Subreddits adjacent to your niche are the highest-signal source for unfiltered customer language. Pain points, vocabulary, objections. The titles of upvoted posts are draft article titles.
5. Search autocomplete — the long-tail factory
Type your seed keyword + every letter of the alphabet. Each autocomplete suggestion is a real query someone is typing. Most are too narrow individually but cluster nicely into pillar articles.
The library, not the spreadsheet
Treating these five sources as ad-hoc inputs – spreadsheet with three columns, manual updates – is what makes the calendar fail. The right shape is a library: persistent storage per site, with each topic carrying its source, search intent, projected difficulty, related keywords, and competitor snapshots.
Once it’s a library, the question stops being “what should we write?” and becomes “which library entry should we promote to the queue this week?” That’s a triage problem, not a creativity problem. Triage scales.
Per-site, not per-account
If you run multiple sites, scope the library per site. The kitchen-appliance site’s topic candidates are not the cycling-gear site’s topic candidates. We made this mistake early — published a kitchen-themed outline on a fitness site because the keyword list was shared. Don’t.
WP Publisher’s topic library is per-site by default. SERP snapshots, GSC imports, Reddit mining, PAA, autocomplete — all captured against the site you ran them on, never bleeding across.
The Sunday-night cadence that survives
Once supply is solved, the calendar runs on a 30-minute weekly rhythm:
- Open each site’s topic library
- Promote 5–7 candidates to the publishing queue
- Run deep research on the highest-priority ones (this happens async)
- Schedule the queue for the week, in each site’s local timezone
- Close the laptop
That’s it. No more “what should we write this week?” meeting. No more 11pm panic.
The behavioral lock-in
Here’s the part most workflow articles don’t mention: the calendar that works is the one you stop checking. The first two weeks you’ll keep peeking — has it published, did it land, is the queue okay. By week four you trust it. By week eight you forget it’s running.
That’s the goal. Not a beautiful calendar — a calendar that operates without you.
We built the topic library and queue for exactly this loop. The Free plan includes 3 deep-research runs a month — enough to feel whether the supply problem is actually fixable for your team. Open the workbench.