The Agency Problem

For many people in the web design industry, design projects have a specific start and end date. The end date specifies when the design (the mockups, code, or custom CMS) will be delivered. After the end date, the engagement is over and both parties move on. This way of working grew out of the print industry and as creative folks migrated over to doing more business on the Web they’ve brought this methodology with them. And it makes sense for print…once the print version is printed there isn’t much left to do except work on something else.

Increasingly, though, social software is showing just how detrimental this sort of engagement is for web design. I dub this the Agency Problem. The agency problem is the problem of doing one-off work in a world in which software is becoming a service that needs constant attention. And that constant attention isn’t just the attention of community managers: it’s the attention of designers as well, who need to constantly refine and rework small changes in the interface based on the emergent behavior of the people using it.

No design survives contact with the user. Once users get their hands on it, they break it and use it in ways the designer couldn’t have imagined. No matter how prescient the designer or design agency there is no way to foresee the social implications of software. In other words, we can design the system in which this complex behavior occurs, but we can’t predict what will happen. It’s increasingly likely we’ll have to make changes to keep the system in equilibrium. But if the agency who designed the software has now moved onto other projects, or the site owners don’t have enough capital to engage with an experienced design group again, the fidelity of the system is in trouble.

So, what can we do about the agency problem?

Well, for one, I think that we’ll see an increasing number of design projects have open-ends to them. This will give the site owners flexibility to grab small amounts of time from the designers to do ongoing design maintenance, fixing smaller interaction problems that crop up over time. And, now that I think about it, this is a growing part of my business. People are realizing that they don’t need huge redesigns…just a lot of small tweaks, from copy-writing to sign-up flows to calls-to-action. If you add up a lot of small, tested & verified changes, the outcome can be pretty substantial. If you do a complete redesign you might be throwing the baby out with the bath water.

Second, I think more social design will be done by in-house designers. My friend Andy Budd (who co-founded the Brighton, England-based design agency Clearleft) tells me that they try to wean their clients off of them slowly by gradually pushing the work in-house. They might help the in-house design team get up to speed, but then they give them the reins going forward. This benefits both parties and seems like a good approach going forward because one-off designs aren’t well suited for software becoming more social with each passing day.