Expert View: How To Implement A Data-Driven Culture

We talk to six of the world's top data leaders about implementing a data-driven culture


We are now deep into the big data revolution, and the majority of organizations have at least recognized that they need data in order to become intelligent. For some, this means buying the next flashy software someone tries to flog them, hiring a data scientist, and leaving the two to get on with it. Put your feet up, fire up the fondue set, and wait for the insights to come pouring in.

Companies that do this - while admittedly likely doing better than those that do nothing - will fail. To use data to its full potential, decision making across the organization needs to be based on data as opposed to gut instinct. A recent study by MIT Sloan Management Review and SAS ‘The Analytics Mandate’ concluded that an ‘analytics culture’ is the driving factor in achieving competitive advantage from data. David Kiron, executive editor for MIT Sloan Management Review, noted: ’We found that in companies with a strong analytics culture, decision-making norms include the use of analytics, even if the results challenge views held by senior management. This differentiates those companies from others, where often management experience overrides insights from data.’

For organizations to instil a data-driven culture, they need to evolve from merely reporting data to inspiring and invoking action. Data must be easily accessed and understood, and organizations must ensure employees have the knowledge to ask the right questions, and the appreciation of its importance to use it.

There are, however, a number of challenges to instilling a data driven culture. A NewVantage Partners’ survey found that over 85% of respondents report that their firms have programs in place to build data-driven cultures, yet just 37.7% report success thus far. Top of the list is getting buy-in from senior management. In a recent survey of 2,165 data professionals commissioned by KPMG and conducted by Forrester Consulting, 49% of respondents said their C-level executives don't fully support their organizations' data and analytics strategies.

We spoke to 6 experienced data professionals from some of the biggest names in the industry about why it is so important to create a data-driven culture and how best to do it.

Joel Shapiro, Executive Director of the Program on Data Analytics at Kellogg's School of Management at Northwestern University:

Anyone looking to implement a data-driven culture needs to think really deeply about what problems they’re trying to solve and what questions they’re trying to answer. If a company doesn’t give sufficient thought to the specific goals they’re trying to accomplish, then they’re going to fail with analytics. You can’t just hire some data scientists and hope that they come up with great insight. The data scientists typically don’t know the business, and the business folks often don’t know the data science. They need to work well together, which means that they both have to know what they’re trying to answer and why.

Garry Ma, Technical Product Analyst at Facebook:

Being data-informed is great, but being 100% data-driven is a bit dangerous depending on how you define data-driven. For example, if you’re a gaming company and you’re selling in-app purchases. If your goal is basically revenue, launching a prompt that prompts your users every ten seconds upselling in-app purchases is great for short term revenue gains, but perhaps not for long term gains because you will eventually lose these users. So it’s important to look at the design and consider the overall strategy of the product.

How do you structure a team and give it purpose? Setting a metric goal will give it focus. Set metric goals like monthly active users (MAU) or daily active users (DAU) - metrics that are easy to understand and don’t require a PhD. Remember that you are communicating these upwards, to investors, and to other product teams as well.

Creating a culture of accountability is also important. When metrics drop or increase, call people out. Give them props for increasing a metric and, without scolding them, make people realise when they make a mistake and the metric drops. A key component here is reviewing your metrics in a weekly meeting. Sit down with the entire team and review the goals to review the changes.

Gustavo Canton, Senior Director of Research at Walmart:

I think we live in a very competitive world and those companies and leaders who adopt a data-driven culture will have an edge in the market. Analytics needs to start from the top, we need leaders that understand the IT challenges, talent and budget requirements that will get the organization.

Stewart Duncan, Director of Data Science at Simply Business:

For us, being data driven is about making sure that our decision makers have the data they need when they need it. It also means they have to actively use it to help them decide what to do next, what not to do, and where to double-down. They are busy people, so the insight has to be easy-to-use and clearly visualized, which makes our choice of analytics tool very important.

We have taken the approach of making as much data available as possible to our decision makers. We don't want them to have to fret about what questions they can ask of the data, or be overly concerned about taking up the data team's time. Sometimes, the best insights come from a whim and we want them to be able to indulge their curiosity. So if a marketer has a hypothesis that IT professionals are more likely to be looking for equipment cover during the Black Friday sales (because they've possibly bought a shiny new laptop), they should be able to check for a correlation from our captured data easily. Our choice of data architecture and tools really helps in this regard - analytics tools that work well on top of scale-out databases like RedShift make it possible for all that granular information to be made available with relatively little overhead.

Adam Dathi, Senior Business Intelligence Consultant at Yieldify:

There were three main problems that we encountered when trying to create a data-driven culture: accessibility of data, the understandability of data and ‘inertia’. Let’s start with accessibility. When I joined the company, we had far less data available and what was available was laborious to get. The root cause was that the data was ‘siloed’, which basically means that we had a lot of disparate data sources that didn’t connect. In order to solve this, our engineers built a data warehouse (using Amazon Redshift), where we now keep all the data together. The BI team then worked with our Operations team to develop a number of processes that enforced the mapping of these databases at the point of creation. This ensured that the data was accessible for processing and analysis. Next, we had an issue with the understandability of the data. The data collected in the BI Warehouse were raw and difficult to use. We circumvented this through the use of custom tables created in Looker. These tables were summarized, enriched and the fields renamed before being presented to the user in comprehensible language. This created useful, understandable data that no longer required an engineering or data background to grasp. Finally, we had inertia. This is a common problem whenever one attempts to implement change to a business. We had the reports and dashboards, we knew that they had value, but we needed to break everyone out of their routine and ensure that they actually began to use it every day. To achieve this we firstly convinced the managers of the value of what we were doing and made data the focus of meetings and performance reviews. Once people understand that they’re being judged by certain metrics by their managers, they become far more engaged in monitoring and understanding them for themselves. We also selected and worked closely with ‘champions’ from within the teams. Champions were our advocates, consultants, and BI/Looker mentors. They would encourage usage from the ground up, advise us on what reports and data points were useful and ran the training sessions. It’s taken a while to break through this inertia, but we’re finally at the point where data is a necessity for our business to function, rather than a nicety.

Walter Storm, Chief Data Scientist at Lockheed Martin

We are at a point where data-driven decisions may still offer companies a competitive advantage, however we are likely 3-5 years out from advanced analytics being table stakes and critical to the viability of a company to even remain in business. That being said, if you want to foster data-driven decisions, it needs to start at the highest levels, where employees see that executives won’t commit to a position without the analysis of alternatives and the data that drives it. Be warned however - it is dreadfully easy to be fooled by this modern-day Merlin.

Big data hype small

Read next:

Is Big Data Still Overhyped?