"That’s what Comma do differently: they connect data and people..."
What’s the biggest lesson you’ve learned in your career?
When it comes to data, the customer isn’t always right. It’s not like buying a coffee, where what the customer wants, the customer should get. In data, the challenges can be complex, and someone who understands data really well will know that it can take the customer time to understand some of the concepts and why they should take one course of action over another. You sometimes need to take customers on a journey of understanding and making sure they are doing the right thing for them is the only way. Stop, and take the time to explain that you are on their side. That’s really central to the whole Comma approach: we’re not going to let you make mistakes.
Why did you join Comma?
I was looking for an organisation driven by values. Comma are an organisation that puts the customer first, respects their own staff and does the right thing. These aren’t just lofty ideals, they are good business sense. When customers see you’ve got their best interests at heart, you quickly become a trusted advisor, not just a consultant. That all stems from letting them know you’re not just trying to make a quick buck: you’re in it for the long run.
How do you think Comma can make a difference to businesses undergoing data/digital transformation?
It’s our combination of deep domain expertise and values. We don’t put the technology first: we recognise that people and processes play as important a role in delivering success for customers as technology. The technology should just enable the connection between the customer and their data. That’s what Comma do differently: they connect data and people.
There’s a lot of talk around AI, machine learning, data virtualisation etc…but do you think businesses are really ready for these data concepts? What steps are they missing?
Lots of fashions come and go in data, don’t they? It's AI, Machine Learning, Virtualisation now, but a few years ago it was Big Data and Data Lakes. I’m constantly amazed how large, seemingly sensible organisations can be so easily distracted by the latest shiny new concept. Right now, lots of organisations are making sense of all these new innovations and investing in these technologies and approaches, but the ones who are really succeeding are the ones who have got the basics right in the first place.
It might be the less glamorous side, but if you’ve got really strong foundational data management embedded in your business – clear accountability, security, data quality, governance, data glossaries – then you have a platform to launch whatever new insight and innovation your business demands. It really increases your chance of success.
Artificial Intelligence undertaken on bad data isn’t artificial intelligence at all. It means you’re still going to do stupid things, but you’ll do them faster. Get the basics right and you can drive an array of innovations.
What’s the worst data mistake you’ve ever encountered, and what did you learn from it?
Earlier in my career, I was working for a large system integrator on a data project. It was early days, but the big decisions had been made. I looked at the solution they’d chosen and thought, “this isn’t going to work”. It was a data problem and they’d bought data technology, but I could see it wasn’t going to address the core business challenges. I was new on the project and I raised my doubts but was quickly told “no, you’re wrong, it’s fine, don’t worry about it”. So I thought, ok, maybe it’s just me.
The programme was a total disaster. It didn’t meet customer requirements, was completely botched, and we had a very unhappy customer. Standing at the coalface of that and seeing it unfold in front of me, I thought “next time I’m in this position, and I see a data disaster ahead, I’m going to trust my instincts.” So that’s what I’ve done ever since. Like I said earlier, if it’s not the right thing to do, you don’t do it.
On another note, I was working with a police force who were complaining about the quality of their data. It was telling them that every vehicle in the area was an amphibious vehicle, which is pretty unlikely. We discovered that the data was being collected from a form filled in by front-line police officers. One question, car type, had the first alphabetical answer selected as default: ‘amphibious vehicle’. The officers, in a rush, didn’t bother to change it. It was a basic data governance issue. One, if the data isn’t important, don’t bother collecting it. Two, if it is important, don’t leave it as a default. Three, make sure everyone understands why it’s important. That busy front-line officer has very different priorities to some data geek. If you want data to work, you need to bear that in mind in your own organisation.
What is the one piece of advice you’d give to any company starting a data initiative?
My advice is: don’t make it a data initiative. Data has got to be at the core of the entire organisation, not a separate project. The way to deliver success here is to look at the data demands across the entire business and deliver innovation by hooking up to a couple of key initiatives. If you segregate it into a standalone ‘data project’, you’re almost certainly going to fail.
How do you think data can change the world?
Well it already is, isn’t it? I do think we will look back on this era and some of it will seem really arcane. We’re only just leaving the first ‘wild west’ of data, and for a long time, the internet has been unfettered and uncontrolled. Everybody recognises that the internet has brought enormous benefits to the world, but for all of the positives – international communication, global collaboration – there is a darker side. Right now we can see this contrast if we look at controversies in social media, election rigging, disinformation, compared to data being used to search for a Covid vaccine and keeping people connected in lockdown.
The good and ill, it’s all part of it, and it shows just how embedded data has become in all of our lives. But I think when we look back on this time, it will look as dated as when roads didn’t have a speed limit, there were no traffic lights, and you didn’t have to wear a seatbelt. Over the next few decades, societies are going to get a handle on how to use data for good rather than ill.