Competitive analysis: five steps to findings you can actually base a decision on

What is a competitive analysis for?

The fastest test of whether you need one is a single question: who are your competitors, actually? Ask five people on the leadership team individually. In our experience, the answers range from "we don't have any" to ten different names. Neither ignorance nor confidence stops your customers from buying somewhere else.

And the question is broader than it's usually understood. Competition includes every alternative a customer has to your offering, including the indirect ones. Ryanair competes directly with easyJet on the same routes. Indirectly, the same airline competes with the railroad, with Flixbus, and with the customer's own car. Since COVID, it also competes in the business travel segment with Zoom, a provider that has never owned an airplane and still made a large share of the demand disappear.

A competitive analysis exists to make a decision better. That sounds obvious, but it's the distinguishing feature: if the analysis leads to the same decision you would have made anyway, it was busywork.

That's why the question comes first, not the data collection. "Should we enter this segment?" leads to a different analysis than "Can we raise our price by seven percent?" or "Why do we keep losing bids to this one specific provider?" Each of these questions needs little data, and each needs different data.

The typical competitive analysis in a mid-sized company hasn't recognized this problem. It takes shape as a slide deck with one page per competitor, neatly researched. It makes one round through the leadership team. Then someone says "interesting," and the discussion picks up right where it stood before the analysis. The reason is a missing question, not poor work.

Which five steps lead to solid statements?

The sequence carries more weight than the method. Every step reduces the material instead of expanding it. That's the real difference from a collection.

Two to four weeks is realistic for an analysis like this if someone has half a day a week for it. What takes longer is almost always step two: the circle gets drawn too wide, and the data collection grows with every additional name.

Where does the data come from when competitors publish nothing?

Among German mid-sized companies, that's the norm. Most relevant competitors are private companies with no press work, no investor presentation, and a website that hasn't changed in years. Even so, the most useful data is closer than it looks.

• Your own lost bids. The strongest source there is, and the one least often analyzed. Sort your last twenty lost bids by who won and on what grounds, and you'll have a sharper picture than any market study gives you.
• Your own salespeople. At every meeting they hear what the competition offers and what it promised. That knowledge exists, but it's recorded nowhere. It takes half an hour per person and a structure.
• Annual financial statements in the Bundesanzeiger. Disclosure requirements mean that revenue magnitude, balance sheet total, headcount, and often the earnings position are public. Two years side by side show the direction, and the direction usually says more than the absolute number.
• Job postings and certifications. A company hiring three people for a new technology is building something. A company obtaining an approval wants into a regulated market. Signals like these appear months before the product does.

What's off limits is just as clear: no posing under a false name, no fake inquiries, no conversations with a competitor's employees about internal matters. It's risky and practically useless. You can't cite information from sources like that as a basis in a leadership meeting, and what you can't cite won't carry a decision.

Which points can actually be compared in a meaningful way?

What's comparable is what means the same thing for everyone involved. That's less than it seems: revenue per employee says nothing about performance when vertical integration differs, and market shares among mid-sized companies are usually estimates, which makes them a gut feeling expressed as a percentage.

Four levels hold up. What the competitor offers and what it doesn't. Which customer it's built for, since almost everything else hangs on that. What price range it operates in, in bands and without decimal places. And how it sells, because the sales channel reveals the cost structure.

One level up, it helps to ask about the forces in the industry: how easily can a new provider enter, how much leverage do customers and suppliers have in price negotiations, what fallback does the customer have. This view goes back to Michael Porter and is useful where the issue is less a single competitor than the question of whether there's any money to be made in this industry at all.

And one level gets forgotten almost every time: the fallback. In many mid-market markets, the most common "competitor" isn't a provider at all. It's the customer's decision to keep doing it in-house, or to do nothing. If you only compare providers, your biggest rival isn't in the table.

What do you do with the result?

In StrategyFrame®, this result moves on to the next step instead of into the archive. The Competition module in the ANALYZE step is where you position yourself, and in the FOCUS step it becomes the value curve: the same attributes you compare yourself on here are the ones you later raise, reduce, or eliminate entirely.

Three to five statements, each with a consequence. No leadership team can keep more than that straight, and anything less doesn't justify the effort.

It helps to sort each statement into one of three buckets. It confirms what you assumed, in which case it costs nothing but buys certainty. It contradicts an assumption that an ongoing decision rests on, in which case it's the most valuable thing in the entire project. Or it's new and unrelated to the current question, in which case it belongs in the file, not in the discussion.

The statements that contradict are the real payoff. They're also the ones that get talked away fastest in leadership meetings, because they're uncomfortable. That's why every statement belongs on the same line as its source: a statement backed by "three of five lost bids, July review" holds up longer than one backed by nothing but an impression.

What makes any analysis worthless?

Trying to be complete. A matrix with twenty providers and thirty criteria has a justification for every cell, which is exactly why it justifies nothing. It also creates a false precision that nobody questions, because the table looks so thorough.

The second mistake is the mirror comparison: you compare on the criteria where you are strong. The result is predictable, and it doesn't explain why customers still buy from the other side. The better approach is the reverse: compare on the criteria the customer actually checks when buying, even if you come off worse.

The third is timing: a competitive analysis that lands after the decision is a justification. If the decision is six weeks out, a lean analysis in four weeks is worth more than a thorough one in four months.