July 13, 2004

allmusic.com's bad redesign

The AllMusic relaunch is the worst I've seen in maybe five years. The amount of visual clutter seems to increase with every page, there's nonstandard DHTML that only works in IE, s a largely-useless Flash navigation widget, an enormous banner ad floating in whitespace, information that used to be in one place has split up into multiple screens, etc. I could go on, but it's sufficient to say any general-public site redesign that requires a manual is a failure on a number of levels. And that's not even addressing the terrible performance problems, both on the browser end and--judging from the frequent unavailability of the site--from the server end.

But this is not the fault of producers who crossed the "should-can line" when specifying fancy gadgets (a term Molly uses when talking about fashion: just because you can doesn't mean you should), or designers who designed interfaces that could only work with IE, or engineers who didn't load test the new design. No. They were never given guidance that would have allowed them to prioritize those things appropriately. The fault lies squarely with management.

A site rollout this bad and a design this convoluted points to a fundamentally damaged management structure. Only bad management, myopic, arrogant or dangerously negligent management can produce something this bad. The site owners took a site with great contents and a loyal user base and, with little regard of those people actually valued, decided to throw it away.

Here's my prediction of how the management decisions embodied in this design will affect the company in the next 24 months:

The management is probably too proud to admit that all of the effort of this relaunch is a failure. They won't throw it out, reinstall the old site and go back to the drawing board to examine the problems that led to this redesign. Instead, I predict that they'll try to push through and "fix" the site, which is of course not what's at fault, really. It's a symptom of a broken process. Someone's head will roll--either the head designer, engineer or product director, whoever "owns" the product. More pressure will be put on the development team, who are probably already demoralized and exhausted from the death march of having to launch the thing in the first place. Any system this broken will certainly have been a death march--the announcements over the last couple of weeks that it was launching were probably as much public threats to force the developers to launch it as they were to inform the audience (how did they expect the audience to prepare for a new design? Such announcements do not build buzz, they're symptoms of corporate fear). Within six months the good people will leave, since they don't need the stress brought on by management panic and pressure (and besides, they can put it on their resume now). New hires will have to learn the system and deal with the residual staff, who are likely to be the least qualified people on the old team, but who now have seniority and will have been promoted into the positions of the good people who left.

Ad revenue for this quarter, maybe the next two quarters, will go up as the sales team can use the new design to extract more money from their connections. The management may even convince themselves that this means it's a success, if a painful one. It's not. Within a year, the novelty of the new design will have worn off for ad buyers, who will be buying less. Sales will put pressure on product managers to create new features that will attract users.

All the while users will be abandoning it. If not actively spreading the word about how bad the site is, they'll not be telling others how good it will be, thus diminishing the quality of whatever brand value AllMusic may have had (the other thing the new site was probably set up to do). At some point, probably two years from now, the company will have laid off everyone, and kept a skeleton crew of editors and writers as freelancers to preserve the illusion of fresh content. This will further dilute the value of the site, and eventually even the tiny cost of maintaining this group of underpaid creatives will no longer be justifiable. They'll then shut it down and sell the assets (i.e. the database of reviews) to someone at a rock bottom price, who will put up the static content with Google contextual ads pointing to Amazon and Ebay. They'll extract the last little bit of value from it, like glue from a racehorse.

I certainly hope I'm wrong--I clearly like AllMusic enough to write all this--but I've seen it happen often enough. So what to do? Here's my first order recommendation to the management of AllMusic: roll back the current design, admit you have a big organization problem that allowed it to happen, try to understand what that problem was, figure out what problems the redesign was trying to address--no the REAL problems, those that you're thinking now are your guesses at solutions to deeper problems, go one level, two levels back, THOSE problems. Meanwhile, try to understand why people come to your site, why advertisers give you money, and start making small changes, one at a time, one per month and watch the effect they have on how people use your site. In two years, you'll probably be making a lot more money than even if all of the gimmicks on the current site worked, which they don't.

Posted by mikek at July 13, 2004 04:52 PM | TrackBack

Excellent analysis, with only one flaw... Most of AMG's business model is licensing their content to other businesses, which has proven hugely successful for them in the past. The question: why bother with a consumer site at all, that competes against their own customers/licensees? Very strange.

Posted by: Andy Baio at July 16, 2004 06:24 PM

thanks for this post mike.

i've been a visitor since the early days, returning 5x a week to research jazz. it wasn't attractive before and the javascript links that you couldn't open in a new window were a bummer, but it met some basic expectations.

i have visited several times in the past week. the first 10 minutes was sad realization. the follow-up visits were just to make sure it wasn't a joke.

Posted by: wayne at July 17, 2004 12:08 AM

FYI: Jeff Veen points out that mp3.com currently licenses all the AMG content and has a much more reasonable interface to it. As per what Andy said, this is probably how it should be: if they make most of their money off of licensing, they should let others compete on the best interface (aka "adding value") to the content and it's unclear why they invest resources in having their own UI (again, I claim myopic or absentee management). But they do. Poor them.

Posted by: Mike at July 17, 2004 03:26 AM

i've been going to allmusic.com on a daily basis for years now (music and information nut), but within one week of this horrible redesign, i've found myself avoiding the site entirely to look for the same informaiton elsewhere.

i had been excited to see the redesign, never thinking it could be this bad and affect the functionality so adversely. the only reasoning i can come up with as an excuse is added pages = added revenue. the more pages they can sell ads on the better. additionally, they may have thought that more pages = longer user time on the site. now it takes 5 clicks to get information that used to take 2-3. very sad.

Posted by: lindsay at July 17, 2004 10:39 AM

I just need to encourage EVERYONE to contact AMG and let them know how we feel about how they have ruined the best site on the web. Please provide them with specific feedback on how to make the site better (bring back the old site sums it up nicely but probably wont work) and let them know that you wont be surfing there until its fixed. I fear that the busines model as described above is too big a hurdle to overcome, but perhaps we should be letting barnes and noble that this might hurt sales that would have come from discovering new bands at AMG?

Posted by: nipsey russell at July 18, 2004 02:50 PM

Zoinks! They BLEW it....

I hope they're not charging their clients too much. What was THE perfect website, is now 5 years below web standards.

Posted by: Roy Estrada at July 27, 2004 01:17 PM

The new site is a complete letdown,
I use it to research artists and their catalogues. In the last two weeks i have not been able to get past search, before I get an error msg. Are there any similar sites that actually work?

Posted by: janine at August 2, 2004 08:54 AM

I can't even get in to the site... It just keeps reloading, and reloading, and reloading...

Posted by: Blomman at August 4, 2004 03:50 AM

The new design is disgusting. I've been a daily visitor to this site for years too, but since the new design, I don't bother. I really liked the way you could find out a lot of information on an artist on one page without tabbing through 15 sub categories... Horrible redesign, the IT geeks have ruined a nice site that used to be easy to access... I'll definately be emailing them. I just wish there was another site around with as big and complete a database..... This sucks!

Posted by: Ventolin at August 20, 2004 12:15 PM

A great example of a situation where a "free" project would better serve the users in question (that is, the ones that want a public music encyclopedia) than a commercial project.

As I mentioned on my blog a few days ago, wikipedia has the right idea (http://www.antisleep.com/archives/2004-08-20_0279.php ). The users control the content; nobody ever "owns" it so it can't be locked up.

Futher, in an internet world, sure, we want a web front end to this data, but we also want APIs to it, and we want to create new ad hoc APIs as necessity dictates. Why shouldn't The GodFather or Tag&Rename be able to programatically access this rich data set? The poster above who said "boo hoo, this stuff doesn't exist to help pirates" is way off base. (a) I buy lots of music but I still like being able to retag my mp3's, add album art, etc. (b) Information wants to be free, and all that. The internet is screaming forward with this stuff, and "boo hoo" arguments can't stop it. Best to harness all that energy and power and figure out the cool new things we can do. I mean that, even as a struggling musician.

If Tower Records or whoever wants a Fuze database in their store, fine, let them pay a company. But consumers know what they want, so why not create a database that consumers control? Sure, it'll take a while for the entire Mingus discography to be entered, but trust me, it'll make it.

Posted by: Scott Evans at August 23, 2004 12:57 PM

Yeah, it would be good if there was an open content music database, and I think there are some out there, but AMG has a big "first to market" advantage, in that it's good enough for most people, so people have little incentive to start the process of reviewing all of those records from scratch.

What's needed is for some group to go at it for a while with little support, until they've built up a competitive database that serves as a reasonable kernel for a large community to build off of. That's a tough call.

Posted by: Mike at August 23, 2004 01:16 PM

It's not just reviewing -- it's getting all the metadata in there. The bio and discography for each artist, the "credits" section for each record. It's a lot of work for sure.

It'd be ideal if existing reviews could be reused on a site like this; not out of the question (depending on the source) and I think even AMG does that.

I really believe that with sufficient server-side resources, this idea could fly. It worked with freedb even though cddb was out there. Or maybe I'm being overoptimisic. :)

Posted by: Scott Evans at August 23, 2004 01:49 PM

The "new and improved" site is a tremendous failure. As an audiophile, I often used the search for songs to identify artists and this often lead to purchases that made $$$ for the industry. This new site is a waste of my time. Whenever I do a search, it pulls up the possible songs, the probabilities and the number of occurrences - and I cannot go any further to identify the artists in those occurrences!!

On my first visit, only about 20% of links were live. When I told this to AMG, I was told it was MY security level set for java scripts. I tried on two other systems with different security settings and they also failed. I gave up and forgot about the site until last night when I foolishly went the again. Now, by some unknown act, areas of the site that did not work before suddenly were functional and I had hope that I could once again resume my search activities - again, I could go no further than the proability and occurrence screen.

Too bad. This was the best site for my musical habits and their improvements have ruined the enjoyment and functionality. I hope they read the comments on this site and roll-back the old functional site until they can repair, test and verify the new look has functionality, not just flash and glimmer.

Posted by: John at August 28, 2004 07:57 AM

When I needed to find out about an artist or thier music A.M.Had no equal now you have gone and dropped the ball.DAMN!!!

Posted by: kenny c at October 12, 2004 07:51 AM
Post a comment

Remember personal info?