Dropbox: "I Hear you, But I’m Not Listening…"

by James Matthews

An observation that has frequently surfaced in our Launching Technology Ventures course centers on an apparent inconsistency in the behavior of some start-ups, who claim to espouse the principals of User Centered Design but do not always practice what they preach.

I will single out Dropbox as the poster-child of this observation. For those of you not familiar with it, Dropbox is a service that allows you to automatically sync a folder of files online and across your computers. Dropbox pays a great deal of attention to its users. Externally, a prominent feature on their website is the Votebox, where users can request and vote on the features that they would like to see added to the product. Having heard Drew Houston speak, I can also tell you that internally a significant portion of their attention is dedicated to working out how users interact with their product. They do focus groups, they analyze usage data – all the usual elements of the feedback loop a lean start-up thrives on.

The alleged inconsistency comes from their observable actions given the feedback we know they are getting. A quick look at the Votebox reveals that the number one feature request is to be able to sync additional folders that reside outside of the ‘My Dropbox’ folder. It currently has 89,722 votes, 60% more than the next highest request, yet has been sitting unimplemented for over a year. What’s going on? Could there be any clearer user feedback around which to center your design? Why aren’t Dropbox listening? Clearly there are users who are frustrated, a comment on another popular feature suggestion says, “This has been in the pipeline since 2 Jan 2010, it has 55,972 votes and yet an item with only 275 votes that was created after was worked on and completed before this was even worked on? So does the "Votebox lets you decide what we work on next!" on or not? (sic) I am wondering now......” Quite.

To me, if you look a little deeper, there is no inconsistency here. The distinction we need to draw is that there’s a big difference between ‘listening to your users’ and ‘doing what your users say they want’. The trick to good UCD is to understand users’ motivations – what are they trying to achieve? What problem do they have that they need solving? Only with this information can you design the best solution to integrate into your product – something you are (hopefully) much better qualified to do than the users themselves. This is especially important for products that strive to be simple. The vocal few who demand the latest features are suggesting the most obvious solution to some problem that they have. They are unlikely to consider the rest of the user base and how they might be affected by the extra complexity that will be introduced. As a manager directing a product’s development, by understanding the problem rather than focusing on the suggested solution you give yourselves the best chance of making good design decisions.

Let us play this out with our Dropbox example. Users are requesting that they can select additional folders to sync. The product designers at Dropbox know that this would create all sorts of problems for some users. For example, an unholy mess could be created if you have two folders with the same name on two different computers with two different sets of contents. Ok, so what are all these tens of thousands of users trying to do? Having not interviewed any of them I can only guess, but one possibility is that many of them want to sync media folders that are in fixed location on your hard drive – their iTunes collection or their Picasa photo store. Perhaps this could be solved by allowing Dropbox to understand these media types and their meta-data and keep these libraries in sync without having to let users deal with the underlying files. Whatever the underlying cause of the feature request, I’m certain Dropbox has asked the question, and I’m certain that they’re working on the solution.