It seems it would be helpful to elaborate on my earlier points about acting like a startup in an established company, and one point in particular. I said “you can’t just take the idea of  a startup and drop it into an established business”, but it looks like that was interpreted too broadly.

A couple of things happened last week around this. One is that Michael Brunton-Spall and I had a long and winding conversation about innovation, in which he claimed to disagree with me about being able to take useful lessons from startups. The other was a Twitter exchange in which Dan Catt was mystified — as was Martin Belam — by Nieman Journalism Lab’s excitement over the Washington Post’s dashboard of usage stats. In that exchange Dan said that actively responding to usage stats “startup 101” while Michael (again) said “The concept of applying startup 101 to news [i.e. our established business] is considered a mistake by some” linking to my earlier post, and then kindly added “That might be slightly misrepresenting that post a little.”

So let me clarify: I have huge respect for startups, and there are lots of things the rest of us can learn from them. But I do have many, many reservations about the hype that surrounds them — almost all of which comes from people not in startups.

Here are the things I think are true about almost all startups:

1. They are driven by one or two individuals who have had a vision, and are risking their own money to see that vision become a reality;
2. There is a small number of people in the company;
3. They do not have a proven business model.

And, in order, here are some implications of those things:

1. The company is run by handful of people who act like their livelihood depends on the company’s success. Because it does.
2. The company is so small that the bosses’ behaviour determines the way everything in the company happens, and they can have direct influence in everything. Similarly, each employee of a startup has a significance which is far greater than if they worked in 1,000- or 10,000-strong company.
3. They need to constantly shift their offering until they have what looks like a proven business model. After that point they’re more-or-less no longer a startup — they’re a company with an established business model they have to execute.

There’s a huge amount to admire in a bunch of people who are in that situation, particularly if they seem to be making a success of it, but I hope it’s clear that you can’t drop all that stuff into an established company. Here is where things are (generally) different in an established company, again point-by-point, in order:

1. The company is run people who have salaries, and who most likely inherited their situation from a previous incumbent.
2. The company is large enough that any one person’s actions have very limited effect. Real change can mostly be brought about only through the co-ordinated action of many people.
3. The company has an established business model with real revenue and the organisational structure is designed around that.

It’s for these reasons an established company can’t just “be like a startup”. You might as well ask for Manchester United to be like a rock band. Sure there are lots they can learn from each other, but fundamentally they’re different.

Also, “like a startup” means different things to different people. I think the original points 1-3 above are the only (near) certainties, but here is a selection of things that other people think means being like a startup. I’ve taken a handful from Jason Goldberg’s excellent “13 Things You Must Do Every Week As A Startup CEO”, which Dan posted in the exchange I discussed earlier:

• Be innovative. Probably not like most startups. Lots of the ones that get lots of press coverage are innovative, of course, because it makes for a great story. But I suspect most startups are just a variation on a theme — as are most successful products.
• Focus on the key metrics, and take action accordingly. Yes, absolutely — but isn’t that obvious? It was Nieman Lab’s excitement over this which caused Dan’s mystification earlier. Surely that’s “startup 101” he said.  I’d go further: that’s not “like a startup”, that’s like any company that wants to take its digital products seriously.
• Separate from any kind of mothership. Well of course they do, because they don’t have a mothership to be otherwise. Doing the same thing in an established company is difficult, though, and not necessarily a route to success. The Telegraph’s Project Euston was intended to focus on innovative ideas and located separately. But it was brought back to the mothership after less than 12 months.
• Blog about your work and your team. Agreed that this is very valuable. But again, not just for startups. Working with social media is important to pretty much any company these days.
• Live and breathe your product, so you know how it feels and can guide it accordingly. Excellent advice. Although a bit difficult if you’re, say, Cisco or Time Warner — what exactly is your product then? Big companies have several products, and probably too many for one person to use meaningfully. A reasonable-sized company will have product teams, and it’s they who should live and breathe their respective products, while the CEO has to trust them (and, of course, use a good number of them).
• Release, release and release again. I really like this philosophy, although not all startups do it — for example, those that are in “stealth mode”.

So startups are great, energised, and fizzing with ideas (because their survival depends on it). There are loads of lessons any of us can take from an organisation like that. But to be “like a startup” is a meaningless for an established organisation, because the fundamentals are different and because it’s debatable that anything more than the fundamentals is actually “like a startup”. Context is all — we should copy the behaviour of others with our eyes open.