Taking this further, ignore what the competitors say they do, but take into account what your market and potential customers do with those competing services.

e.g. A competitor may have launched a bunch of new features that nobody needs or uses. You follow what they do to compete. You then focus on things that don't work when actually, all competitor's customers buy the service for a few core features that they don't shout about, or aren't actually improving (mistakenly).

You can be better at listening to competitors' own customers then they do.

I think this is the spirit of the advice.