Great to meet you, Phillip! I’m looking forward to hearing your thoughts on a data-driven culture. Can you start by giving us a little introduction?
Of course! I’m Phillip Law and I have been working in Analytics for over 10 years, I attended University in Manchester where my undergrad training was in Physics, followed by a Masters and a PhD in Computation Models of the Heart, this is where I really got interested in modelling of complex systems.
The top-level idea behind my PhD was that in the future of healthcare they will be able to image your heart, and map your DNA, and use that info to be able to make a prediction of what drug would be most suitable for that person and hopefully improve treatment outcomes.
Overall a lot of Analysts tend to come from Mathematics degrees, or I find Physics is quite a popular route too! But it’s quite interesting, I have personally noticed a lot of people coming from the social sciences such as Geology or History subjects as well, these candidates tend to make great analysts and are often overlooked when recruiting. But if you think about it this makes sense why they are good analysts, these individuals are able to provide Insight and look at the bigger picture, rather than just being good with numbers; they use the scientific method.
After my PhD, I moved into Finance, which was not for me at all! For example, it was a lot of time spend discussing tactical problems like how to make a programme one nanosecond faster, which in finance can have a big commercial impact. I then took a side step into Web Analytics, and after my first year in the industry I moved over to a job in Adobe, which was really good and was able to work on a lot of interesting stuff and work with interesting clients, it was also quite a technical role. As an Analyst, it’s important to be able to not just pull data out, but understand the technical side of things too.
I then I had a short stint as an Analytics Product Manager and other various roles and then moved to Favourite Story as Head of Insight and have been here for 3 years… which has flown by!
Would you say it’s becoming more vital now for an Analyst to be both Analytical and have the technical skills?
The lines are definitely blurring now. I think that there are two sides to an Analyst, there’s data collection, basically tagging, and then there’s the Analysis and modelling side of it.
What a lot of companies do, and is quite a common mistakes that they make, is they say that want to be data-orientated and data-driven, and then think what that means is that that they need to put all of their data into one place (or database) – and call it the ‘mega brain’ or something like that! They spend years doing this and then once it’s all in one place it’s like ‘now what’.
You then have all the Analysts looking at each other a bit stuck on what to do as the numbers won’t match or make much sense, or they won’t understand it as the Analysts haven’t been involved in the build, so they don’t know what data they’re looking at. I think that’s a big problem that a lot of companies are facing. Start small, fail fast.
Interesting! Can you explain what a data-driven culture is?
Okay! – For me a data-driven culture is more about asking questions and being able to answer those questions with evidence; in this case quantitative data. To me, every meeting starts and ends with the same question…’ What the F*uck do we do now?’. You call a meeting to try to solve a problem, and at the end of the meeting you are then going off to go and solve that same problem.
Having a data-driven culture means if you are going into a meeting with an opinion, and data to back to that opinion. If you get a bunch of people going to that meeting without any data or understanding of how to ask the right questions, you just end up spending a lot of time listening to everyone with no way of knowing what’s right; quite often that meeting will have limited value.
The second part of achieving this kind of data-driven culture is making sure you are getting people close enough to the data so that they can actually self-serve and answer those questions themselves. If a question comes up, they should be able to go away, pull and look at the data and then come to conclusions themselves; it’s making the data accessible.
Thirdly, you need to have the right leadership as well. My experience at different companies is that you could have this massive company, and only a part of it might be data-driven. You can have a hierarchy where someone in the middle of one branch might be data-driven, and what you’ll find is, that everyone underneath will them also be data-driven as that the type of person, used data-driven methodologies and hires other data-driven people. But another part of the business may have more of a go by gut culture. So you really need what’s called apex change, effectively the CEO of the company needs to think and act in a data-driven way to create this data-driven organisation and culture. Every hire and every decision they make, really need to be made by data and feed into that strategy.
It’s really interesting as there have been a lot of talks about creating a data-driven culture. A lot of people think that they need to make sure that everyone has access to the data, and train everyone to have every tool that they have which doesn’t work as too many people have access to it.
Why is it so important now to create a data-driven culture?
That’s an interesting question! I think the thing about data now, is that we are currently in the trough of everyone thinking that data is going to solve all of our problems. Especially in the respects of Machine Learning and Automation – everyone thought that a lot of people would lose their jobs, but in reality, nothing changed. We seem to be moving into a trough of disillusionment.
So that’s why a data-driven culture is so important?
It is now! Everyone needs to have the same goal, and that goal needs to be like a KPI and you need to make decisions based on data and justify what you are doing. If you don’t you’re just being lucky.
I worked at a company a few years ago and worked with a client who thought that they had a data-driven culture, and they were very successful and making a lot of money but within two or three months their whole business changed and started losing a lot of money. Then all of a sudden, within a week or so, their whole strategy changed and instead of going by gut, they were looking at the data to understand what was going on. I think with businesses who are making money, having a data-driven culture is essential because you can make more money by looking at the data if you’re making money without having it! But it’s when things go bad is when you need to look at the data and think why and have this data-driven culture.
If a business owner was thinking of changing their strategy and creating a data-driven culture, what would your tips or advice be?
The first thing would be to make sure you have a data related principle included in the company values. You know how companies have their values and principles on the walls and things like that? – Using data needs to become a value, then you can make sure you are putting data related values into employees objectives so everyone is working to the same goal.
Another thing would be making sure you have the right tools. As a strategy, you can identify each individual team and what data and tools they need and look at each business requirements for each one. You can go to each team and find out what questions they are trying to answer and what data they are going to need to use that and provide training from that.
One of the big things about creating a data-driven culture is data transformation. Data transformation, to some people, can mean almost instantly transforming your company from a caterpillar to a butterfly. But this isn’t the case, it takes a long time. If someone wants to achieve transforming into a data-driven culture it can take years. So that would be another piece of advice; take your time, and also it’s likely to be painful.
I heard with one company who were transforming their company into being data-driven and their aim was making data available to the people. They had screens by the lifts of their office which would provide information and stats about how the company was doing, so when people were waiting for the lift they could take about these analytics!
That same company is a good example of using KPIs to create a data-driven culture. So they wanted to increase the number of people logging into the analytics platform. As soon as they implemented this KPI, they instantly saw an increase in the number of people logging in and using the platform and found that a lot more people were asking questions about how to use it and things like that! It’s the idea of having an aim and a way of seeing if what you are doing is actually working.
What are the signs that a company needs to implement a data-driven culture?
That’s a difficult question! I guess it could be down to something simple like people going into meetings and making decisions not using data to back this and going by gut. Long rambling meetings are another indicator.
It’s one of those things where people are saying that they need to be data-driven but why is it so vital in this age? Is it just because everyone is saying it is. If an organisation isn’t data-driven and still making a lot of money, should they change;
What are the benefits of being data-driven?
It’s a lost opportunity. If you’re not data-driven, you’re 100% losing out on growth potential and opportunity to make more money or doing different things or developing the customer journey and delivering the best customer experience.
I think it was Sky a few years ago who realised that they were providing the same vanilla customer experience to all of their customers and most wouldn’t care about a lot of the packages being offered. So from this, they changed their delivery so they could give a better, tailored and the right customer experience, even the TV ads delivered to your home are personalised to you. That’s a good example of looking at the data and realising you can change something because if you don’t your competitor will!
The data will also be able to give you a good indication of what is working and what isn’t and what you could be wasting your money on. A lot of people say that they don’t know what part of their money is working and what is, and really you should know now, and using data can really help with this, it maybe won’t give you the full picture, but it will at least give you an indication.
I know there are a lot of companies who are still not using the data they have on their customers to their advantage and it’s going to waste. For example, a company who sends spam emails about offers that are irrelevant to their customers is just a wasted opportunity really where they could be using customer information and data to provide tailored offers and improve customer retention. At worst if you aren’t using data you could be hacking off your customer, everyone has bought something and then had a deal come through for a discount on it, it’s annoying.
How should a person manage or lead within a data-driven culture?
That’s a good question! I think the main thing is understanding that there are different types of people. Some people will get some data and they will want to understand every little granular thing and can spend days looking at it. Then there are other people, who will only want to know if they’ve lost or made money and only really care about the numbers, and there’s a massive range in between. As a leader, you have to understand that it’s good to get a range of people. For example, in an analytics team, you’ll want some people who are mathematical or technical and then people who are more analytical. So it’s good to get a blend of these skills.
An interesting way to look at this is by looking at how when they were building Concorde, they looked to hire only the top engineers in the country and it failed! They went massively over budget and spent a lot more money than they should have and couldn’t understand why it was failing. But because they hired only the top aerospace engineers, they couldn’t get anything done, they were arguing and couldn’t make any decisions. If you studied the group of people, you need a range of people; you need the really intelligent person with crazy ideas, and the project manager and someone who is going to get their hands dirty and crack on with the work and someone who’s good at data. You don’t need just the high-flyers with the ideas, you need other skills to contribute to a team to make things work!
When working in a data-driven culture, do you think it’s best to take the problem-solving method or the creative approach to working?
That’s an interesting one! The thing is, if someone came to you with a problem, or say a question like ‘how much money have we made this week?’ and someone goes away to solve a problem. That isn’t solving a problem. A problem is looking and realising that sales are down and then going away and pulling various sets of data, and finding an answer, in a creative way. I guess the answer to your question, would be a bit of both! You have to have a problem to solve and an objective to be able to give it to the team of analysts to solve, and the analysts need to be able to figure out how to do this and be creative in how to solve that problem.
A good example is if you look at campaign analysis. One analyst might look at how much was spent and traffic and how much money was brought in. But its not as simple as that. You can look further and see that someone received an email campaign, which they opened but didn’t buy something until 4 months later so what does that mean and how do you factor that in? Then you need to be creative and realise there is a problem to be solved – so a mix of both!
Great! You’ve had a real variety of roles, can you give us a career highlight?
That’s tough! I’ve done some great stuff working here and have really enjoyed it! Without mentioning any client names, I’ve been involved in taking companies from being not very data-driven and the team not being that analytical to transforming them into being able to change the way they work.
Such as working their data from Excel and then into more advanced software’s and being able to present that data visually on platforms like Tableau. That’s not really a career highlight of a particular moment, but over that past few years, it’s been really interesting and would be a highlight for me.
The best thing about being an analyst is seeing change happen, and that never happens overnight. So seeing that over a long period of time is really amazing actually.
If you were going to start your career again, is there any advice you’d give yourself or do anything differently?
Back when I started, there were a lot of tools around but there wasn’t that many – especially when comparing it to now. You could learn one of two, and once you’ve learnt one or two tools really well, they effectively all work the same. My advice to anyone starting out in their career or if I could start out again would be to pick a tool and learn it really well! I think that will help you in your career as well because a lot of people now are speaking to hiring managers and claiming that they can do Python, R, SQL and various other tools but can only do them to basic or intermediate level – a jack of all trades! Really, if you’re just starting out, learn one really well, to the highest level and then that will get you your first job. Once you have that first job, then you can expand into other skills and tools. Learning one really well will really open a lot of doors for you, more than knowing a lot to a basic level.
Is there anything else that you would want to learn next?
Actually, a bit of a random one! Before Christmas, I was watching and reading about learning to play the piano and for Christmas, someone gave me a piano so that is definitely on my list of skills to pick up as all I can currently play is Chopsticks! People have been saying that if you’re good at analytics you should be good at the piano which doesn’t make any sense whatsoever! Apparently being good with numbers will help, but I’m not too convinced!
Within Analytics, I would like to know a way where you can have an abundance of data on people and how do you find out what are they going to do next. There are a lot of predictive models you can use to do this and companies have scratched the surface when it comes to this, but I’ve not seen in massive detail.
Being able to look at data and sales and be able to see where you need to spend money on each customer. Looking at the customer journey, I want to know, how we can use data to have tailored email for each customer, in sort moving away from that Vanilla experience.
What is it that you like the most about the industry?
Well, to be honest, I used to like that it was always changing and there was something new to learn and new tools to develop. But now, I think there might be too much – but that might be a sign of me getting old! So, that’s what I used to like when I first started out my career; the constant progression and learning new things.
What I love about my job now, is that you can come in every day and there is always something new to do and a new problem you have to try and find the answer to. It keeps me on my toes!
Do you predict any trends for 2019 within the industry?
I’m going to be a bit controversial here, and predict the s the death of the Data Scientist! The thing is a few years ago, there was all of these different data visualisation platforms that kept popping up everywhere, as well as a lot of web analytics platforms. At the minute there are loads of Data Science people, where people are going in a putting all of their data and hiring a team of Data Scientists. But what I’ve noticed at the minute, I’ve noticed a lot of companies popping up that are what you could call ‘Data Scientist is a box’, which they promise to hook up to your data, make a nice GUI and provide a checkbox to get a lot of models running so anyone can use. I think in the next few years, a of these tools are going to become ubiquitous. You’re still going to need a few Data Scientists to understand what’s going on, but you’re not going to need a whole team of 30 Data Scientist to do this. Really, I think the jobs might be automated, as bad as that sounds!
The thing with Data Science, you can make a predictive model, which can be like 80% right, that’s fine and you can get in a Data Scientist which might make it up to 90% right, but what’s the point in that difference – you could just go and make another model? So a bit controversial but that would be my prediction!
Are there any events that you recommend or think the readers should keep an eye on?
Brighton SEO is a good one, that does lots of different analytics events. There are lots of various meetups which I find quite good, there’s a Web Analytics Wednesday one which is good.
There was a conference one I attended called Measure Camp, which are just starting out in the industry which is really good and a free conference and almost impossible to get tickets for! But if you’re just starting out, that’s a really good one to go to, the talks are actually created by the attendees and it’s run in a very unstructured way. We also do our own events here, at Your Favourite Story which are always good and we have one coming up in the next couple of weeks.
To finish, a lot of our readers are going to be recent graduates or starting out in their careers and the industry, do you have any guidance tips to give them?
I would say that if you are starting out or trying to start out a career in Data Science or Data Analytics, try and focus on a problem that you want to solve. Then use R, Python or whichever tool you have chosen as your preferred tool to master to solve that problem.
Instead of watching video tutorials or demos online, pick a project that you want to work on, or problem you want to solve and just try and do it! I think a lot being are really afraid of failing. But the first time you do anything isn’t going to be a great or a success. But just sit down and try a little project or build a website on Python or something, employers like that and like people who have been proactive and done things on their own, rather than just relying on courses!