Why do we do them? To get a sense of the competitive landscape? I.E. how the competition does something better? To understand how to have parity? To understand "white space" in order to innovate? To show standings? To be better informed of customer expectations?
Bleah. So why does it always feel like a waste of time? And why do I dislike doing them so much?
First of all, competitive analysis can only really be performed by an actual customer or user. The actual person who makes one competitor more successful than another. And their standards are informed by the current experience. They like one competitor over another because of a myriad of factors--who they used or encountered first. The content provided by the competitor. If they've "learned" how to interact with the competitor. And ineffable stuff like that they have a personal relationship with a salesperson at one of the competitors.
Secondly, we (UXers) do shitty jobs of doing them. They're based on gut feeling not any real analysis. We apply our biases and we are by no means typical users. We pretend to follow heuristics--aka our expert opinions--but we have no solid methodology for saying one competitor is better with error handling better than another... for example. And we're always wrong. We can't believe that the most successful competitor is better, because it's so ugly. Because it looks outdated. Because it's so Web 1.0. Because we're biased. Because we're fucking arrogant.
Lastly because they set low expectations, and encourage bad design. All the competition sucks. And we want to stay competitive. So we have to stay in the suck so we don't alienate our customers. For example- "everyone else is displaying data in tables. So that means we have to too." Innovation stifled.
And we do them last minute, and when we have our most limited understanding of the client's business model and world, and we don't spend enough time to do them right. They provide shitty data and even worse information.
So let's leave the competitive analyses to the business analysts, to the product managers, to the customers. And instead of competitive analysis how about we spend time coming up with an inspirational landscape of good design that we will reference when trying to solve a client problem?