A competitive analysis is a photograph. It is accurate on the day it is taken and progressively less so afterwards — typically useful for three to six months, less in fast-moving categories.

Monitoring is what keeps it alive. The mistake is thinking monitoring means watching everything, which produces a feed nobody reads.

Monitor changes, not activity

The distinction that makes a watchlist survivable.

Activity is continuous: posts, updates, announcements, mentions. Monitoring activity produces volume proportional to how much your competitors do, which is unbounded and mostly meaningless.

Change is discrete: pricing moved, a tier was renamed, an executive left, a new market appeared in the footer. Changes are rare, and each one means something.

A watchlist built on change produces perhaps a dozen items a quarter, all of which are worth reading. Built on activity, it produces hundreds, and it gets muted within a month.

The signals worth watching

Ordered by signal-to-noise.

High value, low frequency

  • Pricing page changes. The single highest-value signal available. Tier restructures, price moves, features shifting between tiers, and especially a new tier appearing — each is a strategy change made visible before it is announced.
  • Executive changes. Particularly clustered departures, and particularly in finance or engineering.
  • Positioning copy on the homepage. When the headline changes, the strategy changed first.
  • New market or segment language. A footer that grows a new industry page is entering that industry.

High value, moderate frequency

  • Job postings by function. The best leading indicator there is — roles precede shipping by two to four quarters. Watch the ratio, not the count.
  • Changelog cadence. Not each entry; the rate. A changelog that stops matters. So does one that suddenly fills with enterprise features.
  • Review volume and themes. Not individual reviews — a shift in what people complain about.

Low value despite feeling important

  • Social posts. High volume, almost no signal.
  • Press releases. Timed for effect, often announcing something months old.
  • Conference appearances. Bought, not earned.
  • Funding announcements. Worth knowing, but they are the lagging indicator — the hiring that follows tells you more than the amount.

Cadence

Different signals decay at different rates, and matching the check to the rate keeps the workload bounded:

  • Weekly: nothing, for most companies. Weekly monitoring is where watchlists go to die.
  • Monthly: pricing pages, homepage positioning, changelog cadence, job postings. Thirty minutes across a five-competitor set.
  • Quarterly: review themes, community sentiment, leadership, the competitor set itself — new entrants appear and old ones stop mattering. See building a competitor set.
  • Annually: the full analysis, rebuilt rather than updated.

Capture the baseline, or the diffs are worthless

Monitoring only works if you know what changed from what. That requires a record of the previous state.

The minimum viable version: once a quarter, save each competitor's pricing page, homepage and key product pages. Archive services do some of this automatically, imperfectly. Comparing this quarter to last is where the finding lives — a pricing page in isolation tells you what they charge, and only the diff tells you what they decided.

This is also what makes trajectory visible on a positioning map, which is usually more informative than position.

Route the findings somewhere

A watchlist producing observations that go into a document nobody opens has failed in the dissemination phase of the intelligence cycle, which is where most competitive work dies.

Each signal type has a natural destination:

  • Pricing changes → sales, immediately. This is the one that must be same-day.
  • Feature launches → product and battlecard maintenance.
  • Positioning shifts → marketing.
  • Leadership and hiring → the strategic picture, reviewed quarterly rather than reacted to.

The default of “post it in a channel” works only if someone owns reading the channel. Usually nobody does.

Keeping it alive

Watchlists fail predictably: someone sets them up enthusiastically, the volume is high, the signal is low, and within two months nobody looks.

The defences are to monitor few competitors rather than many, watch change rather than activity, run monthly rather than weekly, and give it a named owner with a recurring calendar entry. Unowned monitoring stops within a quarter, without anyone deciding to stop it.

Our Competitive Intelligence report includes a watchlist section identifying which signals matter for the specific competitors analysed, rather than a generic list — because what is worth watching depends on where each competitor is actually likely to move.