For the last couple of years I’ve been running a lean startup within the BBC. No really, I mean it. It is indeed possible to have a successful lean startup inside a large publicly funded corporate. It is all about your outlook.
Journalism Portal: Its about People and Stories
The biggest change for jobbing hacks will be – for those writers, reporters, producers whose job is working on a particular story – the search facility to surface everything we know and have about that story. That is potentially the most valuable element of the Portal. Parallel to that the ability to chat with others working on that story – wherever they are, whatever they’re doing, whichever guests they’re booking.
Mary Hockaday, Head of the BBC Newsroom
My current programme – the BBC’s Journalism Portal – was the subject of an official BBC blog post coinciding with the publication of a white paper on “The BBC Journalism Portal and the Future of Newsroom Production Tools”. The white paper sets out the development and features of the Journalism Portal so far, its uses inside the BBC and some of the challenges we’ve faced while rolling it out. The paper also looks at the future of newsroom production systems and the characteristics of the kinds of platforms that are likely to be needed for journalists of the future.
What the Journalism Portal white paper doesn’t really do is explain my approach to running the programme.
My Lean Startup
I was already the programme manager for Journalism Portal when I discovered Eric Ries’s book “The Lean Startup” (Ries, 2011). What surprised me, when reading the book, was that my approach aligned very well with Eric’s. I was also pleased to see Eric acknowledge that a lean startup is, in fact, possible within a large corporate. Of course I already knew this as that is exactly what I’d been doing for a year and half.
As it happens I’ve already written quite a lot about Journalism Portal on my blog, although not by name. It has appeared in several of my previous posts under the guise of "my current programme". Links to posts about or inspired by Journalism Portal are scattered throughout the discussion below.
In this post I wanted to highlight three aspects of the programme that make Journalism Portal a lean startup:
- Think Big, Start Small
- Experiment
- Pivot
Think Big, Start Small
Eric Ries really pushes the idea of a minimal viable product. The main point is to Build Less, Start Sooner to enable learning as soon as possible.
Journalism Portal did indeed start small. In fact it was live when the programme officially started. The Journalism Portal system began life as a prototype and the programme inherited this and built on it. Admittedly the prototype was ropey and hosted on a couple of borrowed servers. But it was live with real users.
Despite rather modest beginnings there were grand plans for Journalism Portal. The goal was better News planning, better management of resources, more collaboration between Newsroom teams and more efficient workflows.
At the start the system didn’t really do any of that … but it would come. And as we built the capability, we rolled it out, and delivered business benefit on an on-going basis.
I used the same approach – think big, start small – on the infrastructure side of things … which I wrote about in my post on Agile Infrastructure.
Experiment
Eric Ries is also big on experimentation via the “build-measure-learn feedback loop”. Eric basically thinks a lean startup’s main function is to learn via experiments.
On Journalism Portal we learnt a lot. In fact we constantly innovated
One really important point. As the project progressed we simply iterated the site and worked with the portal team to improve, amend and change things – this kind of fast flexibility just isn’t possible with existing systems. The Portal is very adaptable.
Sharon Simcock, Assistant Editor, BBC Radio Shropshire
The goal for Journalism Portal was a flexible and ‘living’ system. Part of that is achieved via the technology but perhaps even more significantly is the approach we are using. The programme exists to drive organisational change within BBC News. Although we knew the general direction of change – open planning and collaboration – the specifics were unclear. And that made an Agile approach ideal.
Very early on I wanted the team to understand the grand visions so we literally built the whole system in a week. This giant prototyping exercise helped us all understand what Journalism Portal was about. During the week we conducted a handful of experiments and these decisively proved the vision of the programme director in areas where others, including myself, had been skeptical. Of course the week also helped the team gel.
Our underlying technical platform is MS Sharepoint. People usually think of Sharepoint as a document repository. What most companies don’t appreciate is that it is very easy to knock up a quick software solution in Sharepoint. That capability meant we used Sharepoint as a rapid prototyping environment. As we went along we developed hundreds of Minimum Releasable Features.
In some cases our quick solutions were part of the strategic direction (e.g. for News Planning), but at other times we’d try a Side Bet to see if some different approach or feature would work within BBC News. Sometimes our experiments failed, and we’d discard the experiment, but when I saw a winner I’d Reinforce Success and build it out.
An incremental rollout helped overcome resistance to change. For example, it enabled us to use Foot in the Door Technique to encourage change.
With all this activity going on I found my approach of Management on the Ground essential to keep across everything. I didn’t want people wasting time writing reports, I wanted them building capability and rolling it out, so I went to look at what they were doing.
Pivot: from Marvellous Monster to Engineered Solution
Unusually for large transformation programmes, Journalism Portal had been successful right from the beginning. This was mainly because the product was live and because my team were very responsive to business demands.
On the downside Journalism Portal was also a giant prototype and although prototypes are great, because they are quick, they are also rough around the edges. After a year of prototyping we’d created what I described as a “marvellous monster”. Great functionality delivering enormous business benefit but very rough and suffering performance and usability problems.
While that was happening our user numbers were growing. And the types of users were changing. We moved from a small population of early adopters to the majority of mainstream users. The early adaptors didn’t mind the rough edges but the main stream users did. So I pivoted.
Eric Ries describes a pivot as decision by the startup to change direction. There are any number of reasons why a pivot might be necessary but in our case it was a “Technology pivot” due to a change in “Customer segment”.
Technology pivot. Sometimes a startup discovers a way to achieve the same solution by using a completely different technology. This is most relevant if the new technology can provide superior price and/or performance to improve competitive posture.
Customer segment pivot. Your product may attract real customers, but not the ones in the original vision. In other words, it solves a real problem, but needs to be positioned for a more appreciative segment, and optimised for that segment.
To appeal to our expanded user base we needed a more performant and user friendly solution. We had to throw away large chunks of our Sharepoint prototype and replace it with an separate application. We moved the product from a marvellous monster to an engineered solution. A massive focus on user experience design plus automated tests (including with Cucumber) were part of the pivot.
This transition took some time and a lot of energy but smoothed out the rough edges and made our main stream users happy. It also made the system much more maintainable.
Final Thoughts
BBC News brought me in to take a different, more Agile, approach to programme delivery. And that is what they got. What they didn’t know is that I was using the same approach I’d use in a startup. A lean startup.
These three elements – thinking big but starting small, constant experimentation, and pivoting at a critical moment – meant that Journalism Portal stayed a business led change programme (with some technology) rather than becoming a technology led project. The latter might have been fun for the developers but was unlikely to deliver the change in working practices, and resulting business benefit, the original business case called for.
There are of course challenges to running a lean startup within a large publicly funded corporate. Bureaucracy. Puzzlement, from outsiders, when groups of people huddle around whiteboards and when there aren’t any obvious gantt charts. And, of course, the normal resistance to change you’ll get in any transformational programme.
I suspect BBC management would have no idea what a lean startup is and would have been nervous that they’d funded such a dubious sounding thing in the first place. But the proof is in the pudding, as they say, and we’ve been delivering clear editorial and financial benefits to the business from day one.
The Journalism Portal system is now firmly entrenched at the centre of BBC News and the organisation is now in the process of deciding the next big strategic direction for the product. There is a good chance that this might entail yet another pivot but Journalism Portal is here to stay.
Thanks
Of course it wasn’t a one man band so a huge thanks to the Journalism Portal crew, past and present, for trying new things and helping the BBC innovate.
References
BBC Internet Blog: BBC Journalism Portal White Paper.
Ries, E. (2011). The Lean Startup: How Constant Innovation Creates Radically Successful Businesses. Portfolio Penguin.