The problem with data: not waving but drowning

When the poet Stevie Smith wrote.....

"I was much further out than you thought; And not waving but drowning."

.....she wasn't talking about the challenge of data overload, but she could have been. In a world where the volume of data surrounding us is accelerating exponentially, senior leaders easily get suckered into a sense that they're doing well because they're gathering and "doing stuff" with all this data, then suddenly they find themselves being dragged under the waves and suffocated as they lose sight of the big picture.

They say the future is in data. And that's partially true, but data in and of itself is not the solution. The most successful businesses ask great questions, gather relevant and useful data, bring it together in a coherent way and then use it intelligently to guide their strategy and action plan.

The problem with data

Today most business leaders would probably describe the use of data in their companies to support their decision making as "in need of work". When your every move in the digital world can be tracked (usually but not always with your explicit permission), the volume of data available to companies can become overwhelming.

Every minute of every day:

• Google conducts 3.6 million searches
• Netflix users stream 69,000 hours of video
• Twitter users send 456,000 tweets
• Instagram users post 46,000 photos
• Youtube users watch 4.1 million videos

From Domo's "Data Never Sleeps" report, 2017

A particular challenge is that most senior business leaders grew up in a small data world and often don't have the skills to be able to translate analysis from big datasets into meaningful output, recommendations and decisions.

As a result they employ data scientists to do it for them, but then the data scientists are often technical specialists and industry generalists, so need guidance to ask the right questions of the data.

So how do you start to get a grip of this challenging problem?

Five steps to use data brilliantly within an organisation

Most organisations are currently working heavily on how they gather, store, extract and analyse data. In terms of scale of task and level of investment required, this is the biggest challenge for sure, especially for businesses with significant and fragmented legacy systems.

However too many companies feel like they're winning because they're doing this bit, especially if they're spending a lot of money to do it.

The problem is, businesses don't win because they have data, they win because they solve customers' problems better than other people. Data can help them do that, but equally many businesses succeed without deep analytical firepower if they have a strong affinity for their customers' needs or frankly simply a great proposition that consumers want to buy into.

To move beyond "I have data, I'm in the game!" to "I use data to drive effective decision-making", businesses need to consider the full end-to-end process for brilliant data use.

1. Good question / hypothesis setting. It all starts here. If you're not clear on what question you're trying to answer then you're not going to get very far. In fact you're going to waste a huge amount of time and effort on data analysis you shouldn't be doing. This is a key step that many businesses are not great at. The lure of getting into the data (and feeling good about yourself for doing so) is too high. This step should be done without having a large pile of data in front of you. It's about questions not answers.
2. Good data access and alignment. You've got to be able to get to the data and you've got to be able to link up different forms of data across different platforms. In one retail business I worked with, the same store had as many as seven different unique identifier codes depending on whether you were looking at the property lease database, the management of retail operations or the store financial performance. Linking those data sources together with one unique code was a simple and hugely value-adding step as for the first time you could see all the data associated with one store in the same place.
3. Good data analysis. Businesses are getting much better at this, especially as they start to build teams of professional data scientists to support the senior team. However analysis is only as good as the quality of the question and the integrity of the data. To my earlier point, data scientists need clear objectives and direction to be efficient and effective in delivering analysis. This doesn't mean that there isn't value in "searching for patterns in data", but the data science team should be clear on the parameters of the task. Analysis should always be statistically validated to prevent variation within the bounds of normality being misinterpreted as actionable insight.
4. Good data interpretation. Data interpretation should rarely be just a desktop exercise. There is of course a role for the unbiased "cold hard review of the numbers", however the nuance of real businesses in the real world often mean that variations in the data can be quickly explained by someone who is close to the detail. This is a careful balance as the data should be allowed to disprove established beliefs that turn out to be incorrect (or no longer correct). Picking the right chart, showing the data in the right way to the right people is a sophisticated art. Almost all businesses generate far too much data reporting without simplicity and clarity to enable senior leaders to make decisions. If you have spent quality time on question / hypothesis setting then this step should be far easier.
5. Good action planning. Making the right decisions, planning, prioritising and executing the right next steps from data is the ultimate key. To do this the data needs to be clear, simple, presented to the right senior leaders in an accessible format and with clear recommendations. Establishing good governance and accountability for decisions and actions is critical.

Tools to help

Many organisations build their own systems in-house, but most are still wedded to Microsoft Excel for data analysis. That's no bad thing if you have a highly skilled team who know how to use it, including the use of automated error checking.

As a business scales, Excel doesn't cut it with large datasets, shared working, more complex analyses. If you have the budget then the big systems providers like Oracle and Microsoft will of course be able to support with systems to manage your data needs. Businesses are also increasingly turning to lower entry pricepoint technologies like Alteryx, Tableau and Polymatica to bring data to life and simplify the generation of actionable insight.

So where to focus?

• Spend much more time on getting good at asking the right questions. Send your senior leaders on training to identify root-cause issues, to challenge established thinking and to frame hypothesis-led problem solving. Review the questions you have set the organisation over recent months and assess whether you have used people's time wisely. The answer is usually to aim for:

"Fewer, Better Questions Asked of the Right People at the Right Time"

• Gather data, it is most definitely the future. Focus particularly on your highest value market and customer segments today and tomorrow. Always be mindful of GDPR and other relevant data protection regulations for data acquisition, storage and use. At the right time, hire some smart people to do smart things with it, but only when you are brilliant at throwing good questions in to the top of the funnel.

"Don't just ask clever people to do clever stuff with no direction, it's rarely an effective use of their cleverness."

• Review the way you interpret data and how you use it to make decisions. Print out every report that gets produced in your business and lay them all out on a table. You'll probably be surprised by how much there is. Work through each report ruthlessly to reduce the volume of output generated and increase the value-add from what does get reviewed. Cut back who gets access to which report to ensure time is used effectively. Establish clear accountabilities for action recommendation and decision making.

"An unnecessary chart is time wasted for the person who built it and also for the person who read it. Use that time more effectively for something else or leave work early to spend more time with the kids."

Not drowning but waving

Follow these simple steps and soon your data strategy will have you feeling confident in the water, able to navigate the waves and the currents to plot your route to the sandy shore.

Photo by Frank McKenna on Unsplash

Gary Crotaz was Group Strategy Director and Customer Director for Mothercare between 2013 and 2018 and is now an independent retail commentator and strategy advisor. If you would like to discuss the themes in this article in more detail then contact him through LinkedIn.