SEO automation software: Ten routines on a rhythm you can see. None of them publish.

No, and that is deliberate. The Scheduled Tasks tab is read-only: no on/off toggle, no cadence picker, on any routine, on any plan. The routines feed each other, so switching the deep crawl off would quietly empty Opportunities, the Topical Map, Issues and the Page Tracker, and you would have no way to connect the two. What you do control is Run now on any routine, the full run history, and the lists that decide how much work there is: which competitors are tracked, which keywords, which AI prompts.

All ten, on every plan, Starter included. That has been true in the scheduler since 11 September 2026, when the last gated routines moved down to Starter. Plans differ by how much you publish, how many sites you run and how many AI assistants are watched, never by which routines you get.

Never on its own. The scheduled routines read: they crawl, score, scan Search Console, check links, track ranks and query AI assistants, and each one finishes by writing data and a worklist. The routines that produce content open a draft or a pull request for a person to approve. There is no path from a schedule to a published page.

Three every day: AI search visibility, the Search Console and Analytics sync, and keyword rank tracking. Three every week: the Search Console opportunity scan, the opportunity worklist rebuild (which also suggests the week's Topic ideas) and the broken link sweep. The deep full-site crawl every two weeks. The competitor authority monitor and the topical map re-scan every month. The year-rollover sweep on 1 January. One hourly dispatcher fires all of them, so there is no per-routine cron to go wrong.

The dispatcher reaps it. Any run that has said nothing for fifteen minutes is marked failed, with the reason on the row, and retried once, resuming from where the dead run stopped rather than repeating the whole pass. The retry is capped at one: two orphans in a row means a person should look, not a third attempt. A failed run stays in your history rather than disappearing.

Yes. Run now sits in the row's menu on every routine, and the result streams in a modal without leaving the tab. On-demand runs are capped per billing period and per routine: ten on Starter, thirty on Growth, a hundred on Professional, two hundred on Agency, with a ten-minute gap between two runs of the same routine on the same site. The deep crawl has its own, tighter numbers, because a large site can take most of an hour to read: 4 a month on Starter, 8 on Growth, 15 on Professional and 30 on Agency, and one per site every six hours. Drafting product cards in the Brand Hub is ten a month on every plan, one per site an hour. The competitor monitor reuses anything less than seven days old rather than re-fetching it.