Name that tune

It's a great idea: a mobile phone service that can identify a song in 20 seconds. But is Shazam a financially viable proposition? Charles Arthur reports

In a bar a week ago, I witnessed what can be described as geek joy, when a mobile phone service worked. Not any service, though. This one identifies tunes when you can't. If you're in a bar and you hear something that you like, you dial a four-figure number (it's not public yet) on your mobile and hold it up to the speaker for 20 seconds. In 15 seconds, you'll get a text message that tells you the name of the song. Last week, those testing it found it identified different Ramones tracks (not easy) and Elvis doing cover versions. You would have to be a music obsessive to do it yourself.

Although the service, called Shazam, has not officially launched, word is spreading fast. The website is at www.shazam.com, and it offers many other things – such as logging all the songs that you've "tagged" on the website, and the chance to buy them, or their associated albums, through the site.

It works by what Shazam call "audio fingerprinting". It claims to have a database of 500,000 digitised song clips, and aims to treble that by the end of the summer. The marketing director, Vijay Solanki, is a former head of marketing at Capital Radio in London, and says: "We're similar to a radio station in that we're a giant repository of music that consumers can engage with."

It's a loose engagement, though, isn't it? Let's examine another claim Solanki made: Shazam will, in the future, let subscribers order concert tickets through their phones. "It's a powerful tool that becomes part of the music supply chain," he said.

Up to a point. It should be a big leap from a clever piece of computing hackery that takes an audio signal, digitises it into a bitstream and compares that against half a million other, stored bitstreams to "becoming part of the music supply chain". At one end, you have delighted geeks saying "Shazam worked! It got the track!"; at the other, disaffected geeks who are not enamoured of the music business, setting up other pieces of hackery to create stuff such as peer-to-peer file-sharing networks and work around copy prevention systems on CDs. What you tend not to find are people saying, "I must use this service so I can buy a concert ticket". They use the phone.

My suspicion is that Shazam is a great idea that fails to do what successful internet- or mobile-based products do: bring people together. Instead, it is driven by marketing people who overlook the reality of it. Think of a few of the internet's hits. E-mail, discussion boards on websites, instant messaging, text messaging, voice over IP, phone calls, even Friends Reunited. They get people in touch with each other, and they are self-building, because anyone can create content for a message board.

No such luck for Shazam. If it doesn't know, you can't tell it what the song was. Shazam has to upgrade its database, a task that won't end. Compare it to www.gracenote.com, the CD tracklisting service, where users enter the details of tracks that aren't on its database.

And there's not much chance of using Shazam to meet other people; chat-up lines that do not work include "Hey! My mobile knows what song this is!" If you're in the position where you're talking to someone and use Shazam to show off, you were already talking, so it hasn't made much difference. Apart from costing you 50p.

In trying to turn Shazam from a clever hack into a money-making proposition, the founders have gone up a dead-end street. It started as "four guys with a plan", and now has 40 employees; I scent dot-com implosion, though I'd be happy to be wrong.

But if it does go bust, someone will surely buy the database and offer the service for free. Perhaps they'll figure out how to let you tell them what it was – the key, in my opinion, to it becoming self-sustaining. It'll still be a brilliant idea. Perhaps then it won't be burdened by people aiming to turn it into part of a musical sausage machine.

www.shazam.com