Call meJeffrey

Yelp.com really bothers me. It’s not so much that it exists, it’s that they make no effort to give one user the ability to seem more credible the next. Their system is so easily exploitable, it ruins the purpose of the site for everyone. Does Yelp not realize this, or simply not care? Either way, here’s some free advice for the makers of Yelp:

1. Have reviewers tag a review as positive or negative before posting. Do a word count, then weight the usefulness of the review. Try to write a useful negative or positive in less than 10 words. “Dude! Best/worst pizza ever, whoa!” is noise, not a useful review.
2. Create a trust network. It’s kinda like a social network, but not lame. Allow users to set how many degrees of separation they trust. Sure, it’ll vastly decrease the content for certain users, but (1) they’ve chosen that so it’s OK, and (2) it’ll add a whole new facet of the site for users who just want to share/discover services with the people they know.
3. Let people quietly mark posts as suspicious. Just having a general “flag this review” doesn’t do anything. It’s entirely too open, so it’s inferred use is “tell us if this is offensive to you”. On a review site, offensive is the least of your worries. Harsh language in a negative review is exactly inversely proportionate to the type of language usually found in a positive review. This is no reason to flag. C’mon, who’s steering the ship? Give your users options so they learn the different flagging reasons you’re looking for. Receiving a flag from a user who’s suspicious of a reviewer using a fake account is way more useful than wasting your time addressing an F-bomb.
4. “Was this review useful/funny/cool?”. Oh, grow up! None of those things matter! Allowing users to review reviews without fixing the credibility issue is spiraling further out of control. Simply put: it’s 100% exploitable. And “funny” or “cool”? You’d get more useful data by inviting your users into your co-lo facility and encouraging them take a dump on your servers.

In the end, the whole point of the site is ruined if it allows people to exploit the most obvious weakness of all: the fake account. Not only am I suspicious of the one shining review in a collection of tarnished ones, I know for a fact that business owners post fake reviews and/or encourage other people to do it for them. So, Yelp… in the interest of your users, the fake account issue needs to be fixed ASAP – but, I’m not giving the solution to that one away for free. Sorry, but I don’t know you.

I was reading the Consumer Reports review on my car out of pure curiosity, and something bothered me. It got low marks for the controls not being designed intuitively. Being a designer, I’m certainly on “team intuitive design”, but that’s because the web is a place where intuitive design really matters. First of all, I beg to differ about my cars controls being counter-intuitive, but that’s just me. Certainly a few things took some seeking out. I’m not seeing this as a big deal. Anything I didn’t get right away, I learned quickly enough. It took me maybe 3 days to fully understand all the ins-and-outs of the car, including the navigation UI. Maybe CR is reporting intellectually to the lowest common denominator consumer.

I’m failing to see the use of reporting on how intuitive the controls are when there’s no realistic expectation of fully understanding the whole car in 30 seconds. If you’re purchasing a car, it’s not beyond reason that you’ll have the vehicle for at least 12 months. The whole point of something being intuitive is so that you understand it right away. This, for example, is very useful for the web because you have casual users coming and going. You make things intuitive so they can get around easily, resulting (hopefully) in a longer stay. In that case, intuition is something to consider.

As I see it, the only way having vehicle controls being counter-intuitive as negative is if you were renting a car. In this case, it could be argued that time spent in the vehicle, much like the time someone would spend on a website, is far more finite therefor it would be a larger issue if you couldn’t just “get in and go”.

For me, the decision to buy a car (or not) usually breaks down to about 75% passion and 25% usability. In many cases passion can (and should) outweigh usability. If a car is sexy enough, who really cares – for example – if the radio volume knob is under the seat. How long would it really take to get used to that? In the 90s, a Saab’s ignition was part of the center console. That counter-intuitive design certainly didn’t result in a shortage of those ugly bastards on the road.

In my humble opinion, information that should be included in the evaluation of a car should only be things that matter in the long run: gas usage, reliability, warranty, service records, etc. Whether or not something is designed intuitively seems to be purely objective. It makes about as much sense as scoring a car lower for being ugly.