The BBC is the world’s largest broadcaster. As such, it sits at the sharp end of the technological revolution upturning the industry in which analogue, sector specific tools are converging with digital, standardised, commodity and Internet-based technology.
In many ways, the BBC has led that revolution. The corporation’s iPlayer service – which allows users to consume BBC content over the web – set the pace for the industry as it wrestled with the move to online video.
But the corporation has also struggled with technology deployments at times. Its 10 year, £1.5 billion engagement with strategic technology partner Siemens has drawn criticism from the Public Accounts Committee in the past. And although the partnership has met the overwhelming majority of its performance targets since, in December 2009 the BBC revealed that it was to move the Digital Media Initiative, an £80 million project to digitise its archive, back in house after Siemens failed to deliver the expected progress. That decision has been credited to John Linwood, a veteran of Microsoft and Yahoo! who was brought in as chief technology officer back in February 2009. His job is to spearhead the corporation’s technology direction, and to reform the company’s outsourcing and supplier management practices. This week, the BBC produced a strategy document, outlining the core principals it plans to uphold in its technology deployments and supplier engagements.
Linwood spoke to Information Age on the eve of the strategy document’s release, explaining the purpose of publishing such a document, what it hopes it will achieve and how technology will shape the corporation’s future.
Information Age – What are the aims of this strategy document?
John Linwood – My role is to make sure the BBC is able to deliver what it is being held accountable for, and the overall aim of the technology strategy is to make the BBC to be more productive. The payback to the license payers, besides cost efficiency, will be an increase in the amount of output we can produce. The strategy is aimed at three audiences. The first is BBC leadership and management, to help them understand the importance of technology. Technology is increasingly part of everything we do at the BBC, and we are getting demands coming from many different quarters – from the audience and from the business – to deliver more and more through technology.
The second audience is BBC staff, particularly the technology staff so that they understand where we are going, but also non-technical staff, because as technology grows increasingly ubiquitous it is important to them too.
Thirdly it is aimed at our vendors. We want them to understand what it is that we are trying to achieve and how we are trying to achieve it. It will provide a way for us to interact with the vendors, in such a way that allows them to give us guidance and ideas, and it helps us guide their product road maps.
And it is a living document. Rather than sitting on a shelf and gathering dust, this will continue to evolve as the requirements change.
Frictionless and standardised
The BBC is always under pressure to provider value to the license payer, and this strategy identifies standardisation of technology as one to deliver that. How will that work?
"We are pushing our suppliers to engineer solutions that don’t require massive change from the user"
The BBC historically has customised and modified technology developed externally, or it has developed technology internally, the reason being it is a large organisation and it believes it is unique in many areas. We often customise things that are not necessarily the right things to customise.
The good side is you get exactly what you want, but the downside is that it costs you more, and it is difficult to support, upgrade enhance. There is a big focus on standardisation of technology across the BBC and creating an environment that encourages people to use standard technologies rather than to customise their own.
One of the core principles of the strategy is ‘frictionless technology’. Can you define what this means?
Technology should be very easy to use, and an enabler for the users of the technology rather than a hindrance. If I ask my wife is she uses interactive services or web-based services on her iPhone, she would say no. But if I ask her if she uses the weather function, she would say yes, absolutely. That’s because it is completely frictionless; she didn’t need to know that she was using an interactive, web-based service. We are now pushing our suppliers and our internal technology people to really start thinking about engineering solutions that are not difficult to use, and that don’t require massive change from the user.
Has usability not always been a priority?
This is more than usability. When somebody plugs in their PC in the BBC and turns it on, it downloads their profile and settings onto their machine, and in some instances that can take three or four minutes. The point is there are 29,000 people taking 3 or 4 minutes every morning just to get their PC to start up. Frictionless technology means getting rid of that delay, so people can instantly access systems. It’s also about not re-engineering the way people work, but flexing the technology to support how people naturally want to work.
The strategy says that “greater interconnection between all of the BBC’s systems … means that technology delivery is becoming more complex and challenging.” Can you provide an example?
Take the example of program scheduling. Five years ago it was a standalone function; you made a schedule and you sent it off to Radio Times and that was the end of it. Now the schedule is connected to our rights management system and our broadcast systems.
And just as internal systems are becoming more interconnected, we are also becoming more interconnected externally, in the way we connect with third party production companies.
I joke that when you change a light bulb in the BBC it ripples through 50 thousand systems.
How will you address this in future?
One way we are addressing the issue of interconnectedness is by building component-based and services-based technology, in which you can essentially replace services without having to replace the whole thing.
For example, we have replaced the production systems at the BBC with digital-based production. Today, editors in the BBC use many different types of edit suite (software), which have different attributes. We shouldn’t have to impose which edit suites they use but equally I don’t want to have to replace our underlying infrastructure every time we roll out a new edit suite. So we’ve architected the infrastructure that allows you to plug in different edit suites.
How does this square with the drive to greater standardisation?
The idea is to provide standardisation, but within that standardisation there must be the right tools for the job.
In what way will you ‘embrace commodity technologies’ as pledged in the strategy document?
Historically, large corporations developed a custom desktop, usually based around Windows. The way it has worked for the past 15 years is that IT departments have controlled and supported that, and they’ve had to adapt and change as new technology comes out. Two things results from this: one is that the IT department is always behind consumer technology – they’re still running on Windows XP while people are running Windows 7 at home. And you also end up with a big support nightmare because every time somebody comes with a new requirement, you need to build a new variation of your custom desktop to meet that requirement.
Imagine if you started at the other end, and told your employees you can bring in a you own client device, and we will make sure that our systems are capable of delivering – in a safe, secure way – all the services and systems that you need to use. This fundamentally changes the role of IT: the user can chose different ways of accessing whatever services that are available.
Of course, there would be conditions; for example, if you plug in a device it has to have anti-virus software.
Suppliers & outsourcing
Like many organisations, the BBC has had some difficulties with suppliers in the past. How does the strategy address supplier management?
"I think historically the BBC even outsourced those people that understood what we were outsourcing"
We are beefing up our supplier management [capability]. It is very important that you understand what it is that you are outsourcing, but I think historically the BBC probably outsourced even those people [that did], so we ended up outsourcing the entire capability to a third party. That makes it very difficult for you to manage the supplier, because you don’t have the expertise internally to know whether you are getting what you are supposed to be getting. And it also makes you a bad customer, in a way, because the supplier can’t engage with you at the level they need to engage.
One of the things I’m doing is beefing up the technology skills of our supplier management capability. The interesting thing is that these are different technology skills you would need if you were delivering something internally. What you don’t want is someone who wants to do it themselves; what you want is someone who can really work through the strategy with the supplier.
And how is your attitude towards outsourcing changing?
We are thinking about what are the right kind of things to outsource, and what to keep in house. We believe that where there is something that is a commodity, like desktop support, another company can do it much more efficiently than us, because they can get economies of scale. Of course, that requires you to standardise, because they can only achieve scale on standard technologies.
A bad thing to outsource would be something that is so unique that the BBC has special expertise in that area, or so core to our business that it represents part of our competitive edge.
Does the Digital Media Initiative fall into that category?
The reason why the DMI should be done in house is that it is more about changing people’s jobs, and how they do their jobs, than it is about technology. As you deploy a change in technology that changes people’s jobs and you make more capability available to them, the way in which they do their job adapts. That isn’t something you can pre-ordain, it is something that just happens.
What we’ve seen with DMI is that the requirements have changed as we’ve learned more about it and we’ve deployed more of the technology. A high change program is probably something you want to "insource".