The problem: rankings are a weekly deadline, not a one-off
Fantasy football content lives or dies on freshness. Rankings, projections and start/sit tools have to reflect injuries, depth-chart moves and snap counts within hours of the news breaking — every week, for five months straight. Doing that by hand means copy-pasting spreadsheets on a Tuesday night and re-checking every number before publishing.
The Fantasy Prophecy LLC solved it the same way we recommend readers solve any recurring content job: build the pipeline once with no-code automation, then let it run on a schedule.
The automation stack
The whole system runs on tools we review here on AutomationPicks — nothing custom-hosted, and no engineering team required.
- Make — the orchestration layer. Scheduled scenarios pull stat feeds, transform the payloads and route them into storage.
- Airtable — the single source of truth for players, weekly stats, projections and tiers.
- An AI step inside Make — normalises messy player names, flags outliers and drafts the plain-English blurb attached to each tier.
- Fillout — collects reader questions and league context that feed our start/sit tooling.
- Scheduled publishing — the finished rankings table is pushed straight to the live site, so nobody has to hit publish at midnight.
How the workflow actually updates rankings
The pipeline runs in four passes, each one a separate Make scenario so a failure in one stage never corrupts the others.
- Ingest: stat and injury feeds are fetched on a schedule and written to Airtable with a run timestamp.
- Score: a projection formula recalculates points per player, then the AI step reviews week-over-week swings and flags anything that moved more than a set threshold for human review.
- Tier: players are bucketed into tiers per position, and blurbs are generated for the ones that changed.
- Publish: the approved table is pushed to the live rankings and start/sit tools, and the change log records what moved and why.
What changed after automating
Weekly production time dropped from most of an evening to a short review pass. More importantly, the rankings update on a predictable cadence instead of whenever there was time — which is exactly what both readers and search engines reward.
The same architecture powers the automated pricing and freshness checks behind AutomationPicks itself: fetch, normalise, review, publish.
See it running
The live product of this pipeline is TheFantasyProphecy.com — the rankings, projections and draft tools published from the workflow described above.
The front end for these projects is designed and built on Lovable.dev, which is how a two-person operation ships and iterates on multiple sites without a development cycle between each change.
Build the same pipeline yourself
Nothing here is fantasy-football specific. Swap the data source and the same four-stage pattern handles product pricing, inventory, lead scoring or any other recurring publishing job.