Challenge · Competitive research
A competitive deck is out of date the day it is delivered
Competitive research is usually run as a project. The research is fine. The format is the problem.
20 August 2026 · 6 minute read
Competitive intelligence fails because it is delivered as a document rather than maintained as a signal. Forty slides, presented once, aging out within a quarter. The fix is to sort competitive signals by how fast they actually change, and to watch each one at its own natural frequency: messaging and positioning shift over quarters, channel mix over months, and share of the conversation over weeks. Almost nothing changes daily, which means a competitor dashboard refreshed hourly is a distraction. Pick five signals, assign each a cadence, and put the output somewhere your team already looks.
Why does the annual audit stop being read?
Because the half-life of its contents varies enormously and the document treats everything as equally durable. The section on a competitor's brand positioning is probably still true a year later. The section on which channels they are pushing is stale within a quarter. The section listing their current campaign was wrong by the time the deck was printed.
Readers learn this quickly. Once you have been burned by acting on a stale slide, you stop trusting the whole document, including the parts that were still accurate.
What actually changes, and how fast?
| Signal | Changes over | Worth checking |
|---|---|---|
| Brand positioning and messaging themes | Years | Twice a year |
| Value propositions and pricing architecture | Quarters | Quarterly |
| Channel mix, where attention goes | Months | Monthly |
| Content themes and campaign cadence | Weeks | Monthly |
| Share of the category conversation | Weeks | Weekly |
| Individual posts and campaigns | Days | Never, on its own |
That last row is the one to internalise. Watching individual competitor posts is the most tempting and least useful competitive activity available. It generates anxiety, it generates copying, and it almost never generates a decision.
How do you set this up without a tool?
- Pick between three and six real competitors. More than six and comparison stops being possible; the exercise becomes a survey.
- For each, capture the durable stuff once: positioning, value propositions, who they appear to be for. A page each, no more.
- Set a monthly check on channel mix. Look at where their last thirty pieces of output went, as a rough split. You are looking for a change in the split, not the absolute numbers.
- Set a weekly check on share of conversation. Count mentions of each brand in your category across the places your customers actually are.
- Write down what would make you act. "If competitor B moves meaningfully into video, we should reconsider Q3" is a useful trigger. "Keep an eye on competitor B" is not.
The most valuable output is the gap
The point of tracking several competitors together is not to see what each is doing. It is to see what none of them are doing. That whitespace is where positioning is cheapest, and it is invisible when you look at competitors one at a time.
What should you do with what you find?
Very little, most months. That is the correct outcome and it is worth saying, because competitive research creates pressure to respond, and responding to a competitor is usually a way of letting them set your agenda.
The decision the research should inform is where to be different, not where to match. If everyone in your category has moved to the same message, that is a reason to move away from it rather than toward it.
The competitive question worth asking monthly is not what are they doing. It is what is nobody doing. The first question produces imitation. The second produces positioning.
If you are an agency
Competitive intelligence is a natural retained deliverable and a poor project one, and pricing it accordingly is better for both sides. A quarterly deep read plus a monthly one-page movement note is more useful to a client than an annual sixty-slide audit, and considerably easier to keep good. It also gives you a standing reason to be in front of the client with something to say that is not a campaign report.
How we do this ourselves
Competitor Radar runs this continuously, reading through the platforms' own interfaces rather than scraping. Per competitor you get messaging themes, channel mix, audience read and a strengths and weaknesses view; across the set you get the comparison and the whitespace. Competitors are suggested from your own brand profile, so the set reflects how you position rather than who you happened to remember.
How Competitor Radar works