How is data shared within your business? Where do models get created and stored? Chances are it is still in Excel. So are new entrants to the data and analytics industry equipped to work with it? David Reed spoke to CACI’s Matt Jarman, Jenny Collins and Laura Jones about the persistence of the spreadsheet that runs the world.
If you talk to students on mathematics and statistics courses about the tools they use, chances are they are working with big data solutions like Hadoop or MongoDB for data storage and R for analytics. Open source software plays an important role in universities, of course, because of the low cost.
Move across to a business environment, however, and you soon find that Excel is the go-to application for everything from creating databases to building models. So what happens when new graduates encounter this entrenched solution? Are they equipped to work with it or is a training crash course required?
Matt Jarman is director of data, insight and visualisation at CACI where he manages a team which handles data, analytics and insight everyday. Laura Jones and Jenny Collins are senior insight consultants, both of them graduates in mathematics from Cardiff University. They talked about their experiences with turning an accountancy-oriented spreadsheet into a cutting-edge analytical tool.
DataIQ: What training did you get on Excel at University?
Matt Jarman (MJ): I did business studies and was self-taught on Microsoft Office at A-level. The university assumed you had some knowledge at a basic level, even in the maths modules. Post graduation, my first job was at Wunderman in the late 1990s - they also assumed you knew Excel.
It is one of the biggest assumptions by employers that everybody knows how to use it. I didn’t get any formal training, even though Excel is important for this work. It is the lowest common denominator across everything we do and is still one of the key delivery tools. We also use Tableau and Qlikview, but because everybody has Microsoft Office it is still and important application within the workplace.
Jenny Collins (JC): In the sixth form, everything was in Word and Powerpoint - I hardly used Excel at school, only very briefly in IT at a basic level. In my maths degree, I did a module within operational research which looked at the coding language, Visual Basic, that sits behind Excel. That was the first time I understood the full capability of Excel.
Laura Jones (LJ): I did the same course and Excel is not taught formally - you have to develop the skills yourself. As a result, the outputs from all the students looked completely different from each other! In the workplace, I have seen people type numbers into cells and then use a calculator to work out the total. So why use Excel?
DataIQ: Are you doing anything to train your teams in Excel?
MJ: We have recognised the differences in experience levels on Excel and Powerpoint and realised the need to run some internal training, if only to reduce the number of queries my team get. Laura is a power user, so it has fallen on her.
LJ: I do some internal training because a lot of people are asking for it. I’ve been asked three times this week alone for help with Excel, like how to change the values on an axis in a graph. It is hard to develop an internal training programme because people’s needs are so different. I have to do some very generic stuff as well as more advanced skills development.
DataIQ: Is Excel still relevant in the world of analytics?
MJ: Excel is an incredible tool, but generic external training courses typically don’t address the specific needs of our business or industry. Clients are often amazed what Excel can do, particularly from a data visualisation and manipulation perspective.
JC: If you can make your outputs look like a web page, your clients know intuitively how they will work. You can build additional functionality with buttons and macros to make it interactive and visually engaging to encourage clients to get more involved in the data.
MJ: We also use Tableau, QlikView, FastStats and a whole range of analytical and visualisation tools, but for our clients, in terms of presentation and delivery of analytics the common link is often still Excel.
DataIQ: Has it got easier to use since it was first created?
JC: Things like the ribbon bar have made it easier, but the Visual Basic coding behind it is still important. There are lots of things you can do that aren’t in that bar, so you still need to be able to code. You probably wouldn’t get taught that unless you studied a maths or IT course, so a lot of users don’t realise it even exists.
MJ: Power users have built their skills around being able to do that. The front end does most of the things most people need day-to-day. But the more complex uses mean you have to get behind that and change things.
DataIQ: Can’t people just Google for answers to what they want to do?
JC: I always like to understand what is going on, otherwise I don’t trust it.
MJ: If you are doing modelling in a lot of new software applications, the front end means users can just drag-and-drop items into a workflow. From an analytics point of view, many users would have no idea how that model actually works. The tool does it for them so they don’t need to. But they may come up with something that is not statistically significant or valid - or they might have just picked the wrong option
LJ: This also applies to the tools themselves. Clients can become quite protective over who is allowed access to them within their organisation. If the wrong person gets hold of them, they may challenge a super-user because what they come up with isn’t consistent with their own outputs. But they may not have realised there are other factors to consider in getting to what the business considers to be a correct answer.
MJ: You could argue that the democratisation of data has made things better - the more people are talking about data and doing things with it, the better it is for people who really understand data, modelling and visualisation. And, in many respects, Excel was the first tool in that democratisation.