Facebook is suffering from an overdose of dog food

The company's staff are so ingrained with the network that they can't make products normal people want to use.

"Eat your own dog food", goes the (rather unpalatable) mantra in Silicon Valley. The argument goes that, without your employees using your own products, how can you know for sure that they are something which your customers would want to use? Take, for example, the fact that Google's Eric Schmidt uses a Blackberry instead of an Android smartphone. How can anyone expect the company to have the kind of top-down commitment to greatness which is so often credited with making Steve Jobs years at Apple so successful, if its own executive chairman thinks another company makes better products? And failing to eat your own dog food is also a symptom, as well as a cause, of decline: consider what the difficulties Microsoft had in dealing with the incursion of iPods rather than Zunes into their employee's pockets said about the eventual victor of the MP3 wars.

But Facebook is in the unusual position of having the opposite problem: its employees use the site too much and too well. In eating their own dog food (and nothing but their own dog food), they end up completely unable to imagine how their typical users interact with the site.

Facebook's site can, if you let it, completely manage your life. You can make an event for everything you do, from work meetings to dinner dates; you can use its messaging client like email, right down to being able to take incoming mail to your address; you can use chat, status updates and photo posting to the exclusion of all else (so long Twitter, Gchat and Instagram); you can check in to your favourite restaurants, follow celebrities and even run a business from your page.

A lot of people use a lot of these functions – but how many outside of Facebook's campus use all of them? And how does that affect the products they make?

Take Facebook Home, a total conversion for your Android smartphone into a Facebook machine. It squeezes all of Facebook's services into your phone's homescreen - whether you like them all or not. Use Facebook just for photos and organising the occasional night out? The chat integration will not be particularly useful for your. What is you prefer to tweet good things you've read rather than post them to your wall? You'll have to change the default settings, then.

But the product which really couldn't exist without an overdose of dog food is Facebook Graph Search. That's a product whose advertised use-case relies on users "liking" their dentist. That might be normal behaviour for the people who Mark Zuckerberg hangs out with, but for most, it's just a bit on the odd side.

The reason why graph search in particular is such a flop is simple: the behaviour it encourages to use it is not the same behaviour it takes to make it work. If you want to get the most out of graph search, you don't have to post a single thing on Facebook – but your friends do. That means that simply launching Graph Search onto the world isn't enough to populate it with useful data, and so it's currently best for finding out awkward facts that people might not have meant to make public

Ultimately, Facebook's major engineering problem for the next phase of its life isn't about trying to make it easier and more fun for people to get information out of the site; it's about trying to encourage them to put it in. But that's a problem that they can't tackle with a company full of employees who are only too happy to live their lives on the social network without any encouragement.

Dog food is an important part of a company's diet – but eat too much of it, and you'll get sick.

Photograph: Getty Images

Alex Hern is a technology reporter for the Guardian. He was formerly staff writer at the New Statesman. You should follow Alex on Twitter.

Photo: Getty Images
Show Hide image

The future of policing is still at risk even after George Osborne's U-Turn

The police have avoided the worst, but crime is changing and they cannot stand still. 

We will have to wait for the unofficial briefings and the ministerial memoirs to understand what role the tragic events in Paris had on the Chancellor’s decision to sustain the police budget in cash terms and increase it overall by the end of the parliament.  Higher projected tax revenues gave the Chancellor a surprising degree of fiscal flexibility, but the atrocities in Paris certainly pushed questions of policing and security to the top of the political agenda. For a police service expecting anything from a 20 to a 30 per cent cut in funding, fears reinforced by the apparent hard line the Chancellor took over the weekend, this reprieve is an almighty relief.  

So, what was announced?  The overall police budget will be protected in real terms (£900 million more in cash terms) up to 2019/20 with the following important caveats.  First, central government grant to forces will be reduced in cash terms by 2019/20, but forces will be able to bid into a new transformation fund designed to finance moves such as greater collaboration between forces.  In other words there is a cash frozen budget (given important assumptions about council tax) eaten away by inflation and therefore requiring further efficiencies and service redesign.

Second, the flat cash budget for forces assumes increases in the police element of the council tax. Here, there is an interesting new flexibility for Police and Crime Commissioners.  One interpretation is that instead of precept increases being capped at 2%, they will be capped at £12 million, although we need further detail to be certain.  This may mean that forces which currently raise relatively small cash amounts from their precept will be able to raise considerably more if Police and Crime Commissioners have the courage to put up taxes.  

With those caveats, however, this is clearly a much better deal for policing than most commentators (myself included) predicted.  There will be less pressure to reduce officer numbers. Neighbourhood policing, previously under real threat, is likely to remain an important component of the policing model in England and Wales.  This is good news.

However, the police service should not use this financial reprieve as an excuse to duck important reforms.  The reforms that the police have already planned should continue, with any savings reinvested in an improved and more effective service.

It would be a retrograde step for candidates in the 2016 PCC elections to start pledging (as I am certain many will) to ‘protect officer numbers’.  We still need to rebalance the police workforce.   We need more staff with the kind of digital skills required to tackle cybercrime.  We need more crime analysts to help deploy police resources more effectively.  Blanket commitments to maintain officer numbers will get in the way of important reforms.

The argument for inter-force collaboration and, indeed, force mergers does not go away. The new top sliced transformation fund is designed in part to facilitate collaboration, but the fact remains that a 43 force structure no longer makes sense in operational or financial terms.

The police still have to adapt to a changing world. Falling levels of traditional crime and the explosion in online crime, particularly fraud and hacking, means we need an entirely different kind of police service.  Many of the pressures the police experience from non-crime demand will not go away. Big cuts to local government funding and the wider criminal justice system mean we need to reorganise the public service frontline to deal with problems such as high reoffending rates, child safeguarding and rising levels of mental illness.

Before yesterday I thought policing faced an existential moment and I stand by that. While the service has now secured significant financial breathing space, it still needs to adapt to an increasingly complex world. 

Rick Muir is director of the Police Foundation