Bumped into a university friend on the train into London this morning, who mentioned that he had met the previous evening with a university friend of his who happened to be an investor in ratedtradesmen.com, so we chatted a bit about why these sorts of sites don't seem to work, when really they should.
His view (just to be clear: my friend's view, not the investor in ratedtradesmen.com) was very clear: people want a personal recommendation, and no website is really going to be able to credibly provide that. A single person saying "here, look at my kitchen, it was installed by so-and-so, you should get him to install your kitchen" is just always going to be a gazillion times more powerful than a hundred ratings on a website which you kind of trust but can't be completely sure that their information is accurate.
Take this site for example: problemsolved.co.uk. Look carefully at the star ratings. Few listings have any reviews, but some listed companies have 5 stars and some have 3, but there is apparently no user data behind this. They tell me they are looking at "the stars issue", but in the meantime they lose credibility. If the stars are not meaningful now (maybe they are meaningful, but it is not obvious what drives 3 vs 5 stars) can a user be sure they will be meaningful "later"?
OR, he (friend on train) said, you need a strong brand. A brand is a reputation. And, really, I think that is the crux of this. Insurance-backed guarantees, OFT-approved codes of practice, user review websites, cowboy-busting directories, etc. etc. are all barking up the wrong tree. Companies (like ours) just need to focus on building a strong reputation and the cowboys will fall by the wayside.
By chance, Jonathan Schwartz (whose blog I try to read because he is probably the most senior business person in the world with a regular blog, but I confess I usually have no idea what he is talking about, all too technical), has this to say today about brands.