I’ve been working at the BBC in the Datalab team for a few months now and thought I’d share a little about what we are up to because (a) I think it is pretty darn interesting and (b) it is a bit different from things I’ve done before due to the ‘interim’ nature of my role.

For the avoidance of doubt I am still working for Notbinary — I’m enjoying being a part of something so new and having the opportunity to help shape the company — and I would never have had this great opportunity at the Beeb without them.

I’m here alongside Mr David Carboni — he is here as interim Tech Lead and I’m here as Product Lead (shocker eh!?). This is a model we are keen on exploring — David and myself (and perhaps a senior UXer or Architect depending on needs) embedding in teams to help with some short-term improvements while getting to know the people better and building up an understanding of what longer term opportunities there are to really fine tune performance. I’m not keen on being ‘body shopped’ or basically being a contractor in all but name so this is a nice way to frame my projects I hope.

So what is the Datalab? The vision I am trying out at the moment looks something like this;

Bring together everything we know about all BBC content in one place and use machine learning to create additional metadata so that we can identify content which is most relevant to individuals’ interests and context.

Build a data ‘platform’, which can be extended by other BBC teams, and which allows many different products to use the data to create more consistent and relevant experiences for our audiences.

Basically it is a multi-disciplinary team but the disciplines are data science, software engineers and data engineers. It is a platform play and our ‘users’ are other BBC teams with audience facing products.

At the moment we are tightly coupled with an app team — reworking an existing product to prove our approach out in the wild. Working on the Google Cloud Platform which team member Beth Anderson brilliant summed up here → https://medium.com/bbc-design-engineering/how-we-deliver-with-gcp-at-the-bbc-1c9812acf3a1. We have been mainly preoccupied with implementing the infrastructure, dealing with the deficits in our data and providing some basic recommendation engines.

To be honest while this has been fun — and I’ve loved being back in the middle of a product team day-to-day — it is the longer term ambitions that are really exciting. The plans for a machine learning platform (operating on a kind of ‘inner source’ model that I have become pretty fascinated with) and the ‘Content Graph’ — which amazingly despite the use of RDF and mention of ‘triples’ I still think sounds great (friends who know my feelings on Linked Data will know how amazing this is!).

Increasingly my personal focus has become trying to provide teams with that ‘psychological safety’ that Google talk about and to give them the space to really become self-organising. This isn’t always an easy thing to do as a consultant as there is something of an expectations of ‘quick fixes’ (I’ve certainly expected this in the past on the other side of the table) but I think a balanced approach supports the long term health of the team best.

Datalab and the app team

The team itself is a real highlight of the project. Young, extremely smart, opinionated but willing to be steered — they really have been a pleasure to work with. My god they make feel old though. I mean I’ve been doing this work almost as long as some of them have been alive 😉

I’ve enjoyed actually facilitating retrospectives and even rolled out the Spotify Healthcheck. The ‘firebreak’ idea I pinched from GDS didn’t really work as well as I hoped but it was nice to have the opportunity to try it out — I totally believe it has a place in the wider cadence of a product team but I need to manage it better in the future.

Not one for this team — but someone who sits nearby is guilty 🙂

Part of my role here as well is helping with the hiring of my permanent replacement. As anyone who reads my blog or has heard me talk in the last couple of years will know I have opinions on hiring practices. I don’t really know why but I expected it to be different — easier — at the BBC. It really hasn’t turned out that way. Despite the brand, the location, a really interesting role, nice offices, great team and a surprisingly brilliant salary (albeit not a public one) it has not been easy to find someone so far. In fact it has been hard work (and the Tech Lead one has been even harder!). The competition for people in London is insane!

Anyway it has been interesting — I tend to talk about product people as ‘humanities’ or ‘(computer) sciences’ and I’m definitely the former and products as ‘platform’ or ‘public facing’ and traditionally I do the latter. This team really needs a ‘(computer) sciences’ style product manager with experience of doing ‘platform’ products. So I have had to do my share of learning on the job!

Also there is a lot to be said for having had an office with air conditioning (and living in a hotel Monday-Thursday with the same) this summer!


Datalab days was originally published in Product for the People on Medium, where people are continuing the conversation by highlighting and responding to this story.

Original source – Product for the People

Comments closed