This is a guest blog post by Rich Pugh, co-founder and Chief Data Scientist, Mango Solutions
Data is the new oil, or so we are told. In some respects, this is true – successful businesses today run on data, and, like oil, data is near-useless unless it is refined and treated in the right way. But refining is a difficult process, and, with many business executives overwhelmed by the “bigness” of modern data, it’s easy to see plug-and-play business intelligence, AI or Machine learning solutions as a one-stop data-to-value machine.
The problem is that all too often, these tools cannot deliver the mythical value expected of them; even if the technology finds an important and relevant correlation, businesses are unsure how to act on the information effectively and understand the full context of the finding. Insight becomes an eye-grabbing statistic in a PowerPoint presentation, or perhaps a one-off decision made based on a nugget of information, and then nothing further. It’s hard to quantify what the long-term value of this was, because the full context is missing.
That’s where data science comes in – or more specifically, a company-wide culture of data science. Rather than just a tool to turn data into insight, data science is a way of blending together technology, data and business awareness to extract value, not just information, from data. While 81% of senior executives interviewed for a recent EY and Nimbus Ninety report agreed that data should be at the heart of all decision-making, just 31% had actually taken the step to restructure their organization to achieve this. That leaves a huge majority of organisations who recognize the potential of data but have yet to find a way to embed a data driven culture within their business.
Restructuring can sound like a difficult and intensive process, but it doesn’t have to be. It’s about following a process to harness existing resources and improve collaboration with a focus around delivering value.
So where do you start? Many companies already have pockets of data science and analytics-savvy professionals dotted around their organisation, but these can be siloed by business function. These can range from product development specialists who understand how to code and develop new analytics solutions, to members of the team who excel at extracting interesting pieces of insight from vast spreadsheets. By connecting these people together into a new Community of Practice – and encouraging ongoing collaboration and connection, as well as discussion around fundamental technologies – you have already created a data science community that sits across your business.
It’s then a case of getting these people to work towards what “best practice” looks like. This requires the team to work together on a common understanding of what the business is trying to achieve, and the questions you want to solve with data, and then build a structure from there for what “good” looks like. As part of this it’s important to agree what the priorities for any projects are, and the ways in which these will be communicated back to others in the business. It’s not about enforcing a one-size-fits-all approach, but instead fostering commonality and cohesion to ensure the team can agree about what needs to happen, when.
Once you have your team of data science experts, it’s time to engage with the business as a whole. Educating the business requires the whole data science team to be confident with what analytics can achieve for the business, and even more importantly, what it cannot achieve that the business might be expecting. This will then need to be communicated in a clear way – using language that the business teams will understand will help break down any preconceptions. This can be daunting, and often, data science teams will find themselves faced with a huge variety of interest levels. Many who hear about the potential of data science will feel it has little bearing on their work – and discussions about its potential will go in one ear and out the other. However, there will also be people who are inspired by what data can do for them and want to get more involved. These people can be future champions for driving a data driven culture beyond the core team.
Most importantly, the business needs to be engaged around relatable, real topics. While the data science team is educating the business, it’s also important to encourage the business to “educate” the data science team. Workshops discussing what success looks like for each business area, what the decisions that shape this success are, and what would be useful for improving those decisions help to transition data science from a magical black box spitting out insights into a process focussed on solving real business issues. From these meetings, the data science community can prioritise and execute around the core challenges that can be addressed with data.
Finally, it’s about finding a way to quantify the value the data science community now brings to the business and make success a repeatable part of the business process. The individuals from the business team who were initially positive about the potential of data science can be fantastic advocates here, explaining in business terms what value a solution has brought – and how these solutions continue to transform business decision making. This can then provide a springboard for targeting more sceptical business departments and scaling a culture of data driven success throughout the organisation.
By adopting a data driven culture, businesses stand a far greater chance of success in the Information Age than by investing in plug-and-play solutions and hoping for the best. By building data science solutions around real business problems, in conjunction with the whole business team, organisations are more likely to see the value thanks to an ongoing culture of problem solving with data science.