Everything your competitors did this week, in Slack by Friday noon

What is the exact prompt?

Set it to run Friday at noon:

• "For each of [competitor A], [competitor B], [competitor C]: read their changelog or release-notes page, their pricing page, their blog, and their careers page. Compare each against what you found last week and report only what CHANGED."
• "Classify every change: SHIPPED (a feature is live), ANNOUNCED (promised but not live, and say so explicitly), PRICING (any change, including repackaging), HIRING (new roles that signal direction, where a first hire in a function counts and a fifth does not), NOISE (rebrands, award posts: one line, grouped)."
• "Write it as an internal newsletter: a two-sentence 'so what' at the top only if something matters this week, then one short section per competitor with a link to every source. If nothing meaningful changed, say exactly that in one paragraph. Do not pad a quiet week into a long one."
• "Post it to [#competitive-intel] and append it to the running Notion page [Competitor log]. Do not editorialize about our roadmap and do not suggest responses. Report only."

Why does SHIPPED versus ANNOUNCED matter?

The classification forces a check. To call something SHIPPED, a companion has to find the tell: a docs page, a changelog entry, a dated release note. Without it a launch tweet becomes a fact about the product, and three months later the feature still says "coming soon". Competitor feature tracking applies the same discipline continuously.

The hiring section is the underrated one. A first enterprise-sales hire, or a first hire with "compliance" in the title, says more about next year than a quarter of blog posts.

Report only, no response recommendations

The moment the newsletter says "we should respond by...", it stops being intel.

Keep the artifact factual and let your team argue about responses in the thread under it.

The running Notion log is the long game: fifty-two weekly snapshots is a competitor history nobody would compile retroactively, and it feeds battlecard work and pricing watches once the raw record exists.

The checks before you trust it

Spot-check the "changed" claims for the first month. The failure to watch for is a page reported as changed because a testimonial rotated. The fix goes in the prompt: "a change is a difference in the substantive content (features, prices, plans, positioning copy) not in layout, testimonials or legal text."

After that every claim carries a source link and your team clicks them. Wider monitoring is competitor monitoring automation.

Frequently asked questions