Turning data into insights

Research shows that many organisations are still focused on simply collecting as much data as they can rather than analysing it to extract meaningful insights. This is a big mistake.

By analysing the data, you should arrive at various insights. Presenting these insights in a helpful way is a key step in turning them into actionable knowledge. I believe the ultimate goal in businesses of any size should be to have data inform decision making across the business – as opposed to gut-based decisions leading the way. To make this a reality, the key decision makers need easy access to data-based insights that are clear and easy to understand. I talk more about getting insights to the key people in your organisation later in the chapter.

How do you help make sure your data leads to insights and actionable knowledge? The answer lies in your strategic business questions, which I cover in detail in Chapters 10 and 11.

It’s important to clearly set out your key strategic business questions before embarking on any data project. Understanding exactly what it is you want to know leads you to better, sharper insights and to the information you need to make decisions. I set out a process for identifying your strategic questions in Chapter 11 but examples of these questions include:

• How should we redesign our website in order to maximise online sales?
• Which customer groups should we focus our marketing efforts on?
• What is the best location for a new store?
• What is the most efficient route for our delivery truck?

Sadly, in business, decisions are often made based on the information available at the time rather than the information that is really needed to make those decisions. By starting with your key strategic questions – defining what it is you really want to know – you can then gather the information needed to answer those questions and make decisions.

Setting out your strategic business questions leads you to the best data. In order to turn that data into insights that help you answer your questions, you need to analyse the data (see Chapters 6 and 9).

Translating insights into actionable knowledge

You know that interesting insights don’t automatically result in action. But, if you set out your key strategic questions in advance, then the insights you gather should help you answer those questions.

The knowledge you gain from answering your strategic questions then needs to be turned into action. You (and your people) need to make decisions based on that knowledge – and then follow those decisions through. If the knowledge gained is not turned into action then the whole thing is pretty much a waste of time and resources.

In their book The Knowing-Doing Gap: How Smart Companies Turn Knowledge into Action, Jeffrey Pfeffer and Robert Sutton (Harvard Business Review Press) explain why many companies fail to turn knowledge into action. They argue the knowing-doing gap (where knowledge is not implemented) is the most menacing phenomenon most organisations face today, costing billions of dollars and leading to a wide array of failures. One reason for this gap is what the authors call the smart talk trap, in which talk becomes a substitute for action and decisions don’t actually result in any changes.

Closing the knowing-doing gap requires a number of changes in an organisation. It’s partly about changing processes and technology so that you can gather and analyse data. But it also requires something of a cultural shift to one that emphasises insights and action. I talk more about shifting to a culture of data-based decision making in Chapter 13.

Here are my top tips for translating insights into action:

• Try not to get distracted by interesting insights that have nothing to do with answering your strategic business questions. There may be scope to revisit those insights in future (they may, for instance, lead to new strategic questions to explore) but, for now, focus on what you set out to achieve.
• Gather all the information you need to answer your strategic goals. Avoid the trap of making decisions based on the information you have to hand. If you don’t have everything you need, see if you can revisit the data capture process to get exactly you what you need.
• Arrive at an answer (or a set of insights) related to each question you set at the start of the process. Depending on your questions, you may need to distribute those insights to key people in your organisation. These insights should be presented in a clear, accessible and easy-to-understand way, not buried in lengthy reports. I talk more about this later in the chapter.
• Together with the relevant people in your organisation, you need to take each of those questions/answers and decide how best to move forward based on the knowledge gained. It’s a good idea to break the actions into specific tasks and milestones so everyone is clear who needs to do what and when.
• Revisit the whole process after a few months (three to six months usually makes sense but you may need to do it sooner in your business) to see if the actions you decided upon have been carried out and led to the results you expected.

Here’s an example of moving from data to insights to action. Sociometric Solutions produces employee name badges with built-in sensors that detect social dynamics in the workplace. The sensors measure how employees move around the workplace, whom they speak to and even the tone of voice they use when communicating. Bank of America used these name badges to collect data on its employees. By analysing this data, Bank of America found that their top performing call centre staff were those that took breaks together. They turned this insight into action by instituting group breaks as standard. As a result, performance improved 23 per cent.

Feeding humans and machines

Data can be fed to your people through dashboards (which I talk about in Chapter 11) or simple reports. The goal is to help them make better decisions that lead to improved business performance.

Or data can feed into any machines that form part of your business operations. Often, this process of feeding data to machines and the machines making decisions based on the data can be automatic, without any human intervention. You can find out more about this later in the chapter.

Grabbing attention

Data can be presented as a number, a written narrative, a table, a graph or a chart. The best approach may actually involve a combination of these formats.

Think about the front page of a newspaper. What makes you stop and pick it up when you actually only came into the shop to pay for petrol or buy milk? First of all there’s the headline: a short, snappy, attention-grabbing description that makes you want to find out more. Then there’s usually a picture that puts the headline into context and adds interest. And underneath there’s a short narrative that introduces the story and gives the key information.

Unfortunately, I rarely see this approach in the material that organisations distribute to their people. There’s usually complicated graphs or charts that no one understands and the key takeaway points are buried in lengthy narrative that few people read all the way through. It might look snazzy or impressive, but the key messages are hidden. And if the messages are hidden, people aren’t likely to act on them.

Try to present your information as a newspaper would. Start with an interesting and informative headline that summarises the main finding. Include a useful visual (such as an image, graph or chart) along with a short narrative that sets out the key messages and supports the visual. Use colour where appropriate, but don’t get lost in style over substance.

Making the insights easy to access and digest

Visuals are great for conveying information because they’re quick and direct, they’re (usually) easy to understand, they’re memorable, and they add interest (being much more likely to hold the reader’s attention than a full page of text).

Narrative is also important because it, without it, people can interpret the data differently. With a short narrative you can ensure everyone understands the data in the same way. It also gives managers a story that they can easily share or filter down as appropriate.

Using visuals and narrative together is much more powerful than using either on its own. For instance, a graph detailing sales history is extremely useful for analysing trends over time, but a narrative can put that information into context – explaining what might be behind that trend.

In terms of access, it’s important that your decision makers can get at the information they need when they need it in order to make the best decisions. It’s no good disseminating monthly reports if key decisions need to be made on a daily or weekly basis. Restaurant chain Dickey’s Barbecue Pit offers an excellent example of getting the right information to the right people at the right time, in the sidebar ‘Big data and barbecue – A surprising flavour combination’.

Understanding machine learning

Machine learning involves feeding data into machines, which then determine the best course of action based on that data.

Machine learning refers to the fast growing field of creating self-learning algorithms that can adapt themselves based on given data without any human intervention. In a nutshell, the machines learn from the data they’re given and decide what to do next.

Sometimes the course of action decided by the machine may need human intervention (for instance, if something needs repairing or replacing). But increasingly computers are able to carry out the intervention themselves. An example of this can be seen in what Rolls-Royce refers to as its Ship Intelligence Initiative – see the nearby sidebar, ‘Rolls-Royce’s space-age initiative’, for an insight into how this works.

Connecting data with machines

The potential for machine learning and data is all very exciting. But how are businesses already using this technology? Here I set out some great examples of how businesses are already connecting their machines with data, allowing access to a whole range of useful insights that we haven’t had before, even just a few years ago.

By connecting data and machines, the insights you gain can help you increase efficiency, improve product quality, provide a better service to your customers, cut costs, make your employees happier and much more.

I should clarify here that you don’t need to be running a manufacturing business with big, expensive machinery to care about connecting data with machines. Machines can, in fact, refer to your website infrastructure, for example, using algorithms to tailor your website based on the visitor’s IP address (perhaps by flashing up an offer on overseas shipping).

It could even refer to office equipment! Humanscale builds sensors into its line of office chairs, standing desks and work stations and offers its OfficeIQ system to monitor workplace activity such as how much time individuals spend sitting or standing at their desks as well as how long they’re away from their desks.

In Ireland, grocery chain Tesco has its warehouse employees wear armbands that track the goods they take from the shelves, distribute tasks, and even forecast completion time for a job. In other sectors, including healthcare and the military, wearables can detect fatigue that could be dangerous to the employee and the job they perform.

I mention Rolls-Royce a few times in this chapter because it’s one of the foremost examples of sensor technology and machine learning in action – and a great example of a relatively traditional company embracing big data. The company is already connecting data with machines in order to improve the way the business operates. Rolls-Royce manufactures enormous engines that generate huge amounts of power as they propel airplanes and ships across skies and oceans. These engines and propulsion systems are all fitted with hundreds of sensors that record every tiny detail about their operations and report any changes in data in real-time to engineers who decide the best course of action. In addition, the company’s manufacturing systems are increasingly becoming networked and communicating with each other. At its factories, the innovation is not just in the way machines bash and shape metal – it’s in the way the company automatically measures aspects of the manufacturing process and monitors quality control of the components being made. At its Singapore factory, Rolls-Royce generates half a terabyte of manufacturing data on each individual fan blade. With 6,000 fan blades being produced each year, that’s three petabytes of data from the manufacture of just one component. And these concepts are creeping into all manner of manufacturing areas: check out the nearby sidebar ‘After smartphones comes smart … shirts’ if you need to be convinced!

Connecting data with processes

Big data is about so much more than how much data you can generate. It’s about analysing that data in order to draw out insights that drive efficiency and progress in your business.

Your processes and systems can also be connected with data, so that you can improve how you do things based on what the data shows.

Big data analytics helps Rolls-Royce identify maintenance actions days or weeks ahead of time, so airlines can schedule the work without passengers experiencing any disruption. This idea can be scaled down and applied to any company with machinery or vehicles that require regular maintenance. So, if you run a delivery fleet, your servicing schedule could be connected with data that monitors wear and tear on the vehicles, enabling you to service or repair vehicles when they need it rather than when an arbitrary timetable says they need it. Maintenance can be scheduled in advance to avoid vehicles being off the road at inconvenient times.

Not only do approaches like this help you run your business in a more efficient way, they also reduce costs, especially when you’re able to fix things before they go wrong.

Connecting data with processes also helps you offer a better service to your customers. For example, monitoring your stock control system helps you ensure items stay in stock at peak ordering times and allows you to run dynamic promotions based on stock levels. It allows you to measure data from your customer service team and make changes that improve your service or product based on what the data tells you. It’s about much more than machinery and metal and cables – it’s about making your business more competitive and securing its future.

Transport for London (TfL) oversees a network of buses, trains, taxis, roads, cycle paths, footpaths and even ferries used by millions every day. Running these vast networks, so integral to so many people’s lives in one of the world’s busiest cities, gives TfL access to huge amounts of data. This is collected through ticketing systems as well as sensors attached to vehicles and traffic signals, surveys and focus groups and of course social media. This data feeds into TfL’s processes in two ways: planning services and providing information to customers.

The data is made anonymous and used to produce maps showing when and where people are traveling, giving a far more accurate overall picture, as well as allowing more granular analysis at the level of individual journeys, than was possible before. This allows TfL to understand load profiles (how crowded a particular bus or tube line is at certain times) and to plan interchanges, minimise walk times and plan other related services, such as retail.

In this chapter I share some very impressive examples of how companies are acting on insights from big data. At present, a lot of this innovation is driven by larger corporations who have the manpower and resources to generate and analyse huge volumes of data. But that doesn’t mean this is all beyond the scope of smaller businesses. In fact, I’d say the opposite is true.

Smaller businesses are often much more nimble than larger corporations, making them better equipped to quickly act upon what data tells them. You may not be generating petabytes of data in your daily operations but that doesn’t mean your data is any less valid. The key is to focus on the insights that the data throws up, communicate those insights to the people who need to know, make informed decisions and follow through with action.

As all areas of business become more data-driven, this process of turning data into action becomes a core factor in success – and that’s true whether you manufacture aeroplane engines or run a small organic grocery shop.