Fire your Developers
Facebook is an incredible tool. It lets you connect with people you haven’t seen for decades, and see how your lives have diverged. But it also helps you find new business contacts, or new friends, or even new love. Facebook, as a company, has succeeded. I read a quote recently (sadly, I cannot remember the source – if you know the source, please let me know so I can give proper credit) that said (approximately) “not being on facebook today is like not having a cell phone – by doing so, you are making a statement, whether you intend to or not”.
They are estimated to have roughly $300 million in revenue. The company is worth somewhere around $6.5 Billion. They have 900+ employees. So why the hell do they keep changing things that people get comfortable with? The front page of facebook has changed significantly at least three times in the past year. Unlike cell phones, where people have an option to select a new one, Facebook has a single primary interface.
Facebook appears to be following the disastrous trend of allowing developers to create their software. No, I don’t mean actually write the software, of course they do that – but create the experience that a customer will see. There are a couple of reasons for this, but none of them relate to the kind of things a $300 million (not to mention $6.5 billion market cap) company should have. Developers are amongst the most necessary people at the inception of a software (or web) company. you can make a product with a developer – it may look like crap, but it’ll work. You can’t do that with anyone else. It’s the same reason QA is always an afterthought – no one ever started a company with a QA person and no developer (ok, ok – outsourced testing companies might – but if you learn about S3′s Production System, you’ll realize that outsourcing QA is a terrible decision, but that’s for another post).
So when you start a company, you hire some developers. And they do a good job. But over time, your product matures. And your developers become less critical. So you have several choices – put them on things that matter to you, but aren’t very technically interesting, fire them, or keep them designing the critical stuff, ignoring all the input you get from your clients (and probably many people within the company) telling you that they’re making bad decisions. Each of these has its downfalls – the first makes the developers bored, and disenfranchised. That sentiment will probably spread through your organization. And then they’ll probably quit anyway. as for the second, firing them feels wrong. These are the people who helped you get off the ground – these people are the reason your company exists. The third makes your clients disenfranchised, and frustrated.
So, if you want to continue to grow, and keep a strong culture, and keep your clients content, this is when you have to fire your developers. It’ll feel bad. And there are days you’ll regret doing it. But if you don’t, you can watch your company erode, and instead of losing some people, you may well lose them all.
Sure, if you happen to be the sort of monopolistic company that Facebook has managed to become, the last option is survivable, but probably only in the short term – some day, someone will figure out a way to overthrow you over time – maybe with just 140 characters.
[...] You know my answer? no, it isn’t “fire people faster” (although I am a fan of that). my answer is nothing. why? because whatever we did has worked. Could we have done something more [...]
People learn and adapt to interfaces even though they may complain on change. Even though the UI has changed and pages has become a bit cluttered at times, I expect it. Google does a wonderful job on covering that up. Maybe Facebook should of done what google does with labs and have people turn on functions instead of forcing them to use them.
Have any of you noticed for normal non-techie people(like my mother and father), security on profiles and personal information is the not as easliy managed as before. Most don’t think about security until it something happens so where Facebook was proactive in security in the past, most people have become retro-active about security. Also the Academia feeling of how exculsive facebook was is gone now. That is what concerns me more.
my point is pretty simple.. metrics like “increased page views” and “more engagement” are good things for the product (facebook’s users are happier and using the product more) and good thing’s for the business (we have more advertising to sell).
just because you don’t immediately like the new UI doesn’t necessarily mean the metrics are wrong or that they weren’t scientific about the design.
And that’s not to say they will get it right every time. sometimes data on a 1% test indicates that things will work well and when you roll it out more broadly, things don’t roll your way (the last design was an example of that).
but to say that you don’t keep iterating because you failed last time, that’s just silly.
Yahoo home page and MSN home page are exactly the same.
And you are absolutely incorrect about Google. They have iterated on their home page and search results page constantly. They’ve also focused on things underneath the cover as well (like response times, etc).
http://cdixon.org/2009/12/17/googles-feature-creep/
And the iPhone’s design has also changed from version to version. The app store didn’t exist in the old version. multi-tasking didn’t exist. etc etc etc.
point is that you can’t stand still.
Seriously??? You think facebook should not attempt to optimize metrics like “increasing page views”, “increasing the amount of time people stay on your site”, “increasing the number of friends people add”, “increasing one’s interactivity with the site (as measured by status updates, comment posting, liking”.
These are all metrics that the home page should be measured against for an app like facebook because they all correspond to improved revenue potential (ie creating greater advertising inventory).
Let’s do a dumb back-of-the-envelope calculation. 350 Million unique users. Let’s say they create a new home page design that yields 1 extra page view per week per unique user. With 5 ad slots on a page view at an average of $1 CPMs.
With those metrics, Facebook just generated $7M of advertising inventory per month (350M unique users * 1 page view per week * 5 ad slots per page view * $1 per thousand impressions * 4 weeks. Now they just have to go sell as much of it as possible.
That seems hardly a bad decision. Furthermore, i seriously doubt that Facebook rolls out a new home page willy nilly.
They probably do what most apps do, roll it out to 1% of the user population and measure it against the metrics above (or any host of other metrics they identify in their data mining that correlate to generating revenue).
The fact that they are changing their home page frequently compared to a company like yahoo or MSN who’s home page never changes is actually a sign that they are operating more effectively.
As a simple example, Google reduced their search result times by 300 milliseconds (think of that.. 0.3 seconds, typically imperceptible) and they correlated that to some crazy multiple number of searches performed per user which of course correlates to more click throughs.
The iPhone is another example. iPhone users perform 50X (no typo!!!) the number of google queries when compared to other smart phone users. The reason- a great killer UI.
These are the types of things that one should absolutely be iterating on, especially if you are an advertising-based company.
Au Contraire, mon frere…
Those are exactly the kind metrics I am condoning.
Just look at the two examples you provided – the first is arguably the best user interface on the web – how different is it from the first day you saw it back in 1998? virtually unchanged. it has a few little bits of text around the outside, but for the most part, it’s still the same text box with two buttons. Talk about consistency – they still have “I feel lucky”. Do you really think that adds any value? a) it drives you away from their revenue stream and b) nobody uses it. But maintaining the consistent user interface that made people comfortable searching with them is paramount to their continued success. Your example is exactly the point – they shaved 300 ms off the average search time – for all intents and purposes, they did not change the customer experience, except by an incredibly small amount, and even then a change that could not be argued as negative.
Your second example is well written about and known as one of the best user interfaces of all time. If you woke up one day, and instead of your iPhone having a 16 (20) icon array, it suddenly had a list, you’d be unimpressed.
You’re attempting to compare the two world leaders of design (and who maintain a consistency over significant periods of time) to Facebook, saying that they’re doing comparable things. But the fact is, they aren’t.
In truth, I believe the last two Facebook changes have been an attempt to undo a terrible decision by a developer for the prior release. My guess on how their history has gone is basically: once upon a time, a facebook developer said “DUDE!!! Have you seen Ajax?? this stuff is awesome!!! We could totally have a feed of everybody’s friends updating in real time” to which someone said “yeah, that sounds great, lets do it!”. they did, then someone said “uh… guys…. seems we can’t keep up with this ridiculous volume, plus all our users are pissed that we changed the UI”. the last two have been trying to undo that change, except they’re missing the most valuable component of a UI, which is customer comfort (you’re still using a QWERTY keyboard, aren’t you?). They have had nothing to do with improving revenue.
If facebook focused its developers on revenue related changes – as determined by real business analysis, I would wager they would not change the primary user interface as much are, or in the way, they have changed it.