Six tips for using existing data more wisely

Six tips for using existing data more wisely

Organisations are collecting more data than ever before and processing them in new ways in an effort to improve their businesses. But often they are not using the data in the right way — if they are using them at all.

Eighty-seven per cent of finance professionals said Big Data holds the potential to change the way business is done, according to the CGMA report From Insight to Impact: Unlocking Opportunities in Big Data. But 86% of those surveyed said their businesses struggle to get valuable information from the data they have. Hurdles to maximising data’s value include organisational silos, challenges related to data quality and an inability to work with unfamiliar, non-financial data.

Here are six ways to make sure you are collecting the right data and making the most of the data you already possess:

1  Consider the broad indicator over the precise measurement. There are many indicators that show us general trends happening in our organisations. But often, organisations spend lots of time and money drilling down too deep into a problem when such drilling isn’t really necessary. Ask yourself: Do you really need to measure it down to the fifth decimal? Or will a more general indicator suffice?

Take employee satisfaction, for example. Does your company need detailed measurements of pay and benefits compared with those of the competition? Does it need to deeply analyse promotion opportunities offered, or the culture inside the organisation, or how good the boss is at motivating and inspiring people — or even the office and business environment itself? You can measure and analyse each of those things, but that could be very expensive.

Instead, you could use a broad indicator such as absenteeism. Generally, if a company sees abnormal spikes in absenteeism, it’s a sign of dissatisfaction. From there, you can dig deeper. It may not be because of any one of the factors listed above; it may, in fact, be something as small as a lack of parking. The benefit of that indicator is that it doesn’t try to explain what’s wrong in your organisation.

In this case, absenteeism is like the “check engine” warning light on the dashboards of many automobiles. It doesn’t tell you exactly what’s wrong, but it tells you that something is wrong and that you should have it checked out. High absenteeism is a good starting-point indicator — one that might prompt an organisation to more deeply examine salaries, culture, managers and so on by doing focus group work and surveys. The best part: Absenteeism data are free.

2  The data that you gather often need to be received more quickly than the impacts and the stresses facing your organisation. Data are often more valuable to you if they are received faster than if they are precisely accurate.

Consider the fast-food drive-through. The cycle time through the drive-through is about three minutes, which means data have to be replenished faster than once every three minutes to be useful to the operations people in that part of the business. On the other hand, the supply chain might work on, say, a one-week cycle. Therefore, the data don’t have to be as fast. Supply-chain managers don’t need to know what’s happening every minute, but they certainly need to know what’s happening every day or every couple of days, whereas setting up a new store location might take a year. The data those managers need have a less frequent drumbeat.

So, even inside the same organisation, you may have different needs for how fast the data must be. But the point is that data need to be arriving ahead of the stresses and the risks that your organisation faces.

3  Make sure leadership is aligned with strategic priorities and sees the same activities the same way. We need a way to establish the most important things we should be looking for and score our performance accordingly. Imagine, for example, your strategy is focused on growth by entering new markets. When you take a look at performance of activities — such as sales into new market areas or the number of new customers you’ve been able to attract outside of your existing areas — those data should have more value than measures that would be important under a different strategy. Determine your focus based on your organisation’s strategy; this will determine whether you look at measures such as the number of customers buying multiple products or the number of repeat customers. Depending on the strategy, consider different measures to ensure alignment on how you look at your organisation’s performance.

In a multi-location business, however, this can be a challenge because you might have one newly established business unit that is growing into the market and another unit that has been there for a long time, and in that business unit, they’re looking for share of wallet. So, you can even have different strategies inside the same corporate entity. You should have visual triggers when leadership’s looking at the data, so they know when they’re stepping over the line from a penetration strategy to a share-of-wallet strategy.

4  Once you’ve determined what the right measures are, be prepared for them to change as your organisation does. Measures can’t remain the same over time.

Think of something like a new product development cycle. In the early part of that cycle — the product idea creation — you might be looking at things like the number of product ideas. During the next phase, you’ll look for product development indicators such as the time from idea to first prototype. The next phase might be product testing, where the measure needs to change to, say, an approval rating from a beta test audience. Finally, during product launch, you might look at sales per month.

5  In most cases, the organisation is already collecting the data you need; you just need to know how to use it. You should always be able to find data within your organisation that indicate how a process is running, whether you’re looking at the inputs, the transformation activities or the outputs.

The main benefit of using available data is that it doesn’t add costs to your organisation. You don’t need more people to gather the information because you’re already gathering it. You don’t need more people to process the information because you’re already processing it.

Secondly, it allows you to get your scorecard and dash - board set up very quickly. You’re not held ransom for data. In some cases, the signal strength of those data might not be as strong as you would like. The quality of data you need depends on the type of decision at issue. You do not need Six Sigma accuracy for all decisions. For instance, the data you need to extend a marketing programme for a month can be less accurate than the data required to terminate an employee or close an office.

Using information that already exists creates higher ownership faster. You’ll gain much more acceptance when you begin using indicators that people in the organisation already know and respect rather than those you have imposed on them.

And because you’re using existing data, your net training cost is significantly lower.

So, it’s often faster, cheaper and better to start with broad indicators. From there, you can determine whether to investigate further, and, if so, where — and how deeply — to probe next.

6  Present the right data in the right way to the right people. Think of an organisation as an airplane that needs to see the business world around it at various heights — from 30,000 feet to ground level. Each altitude level needs its own form of information presentation.

Senior management wants to cruise along at that 30,000-foot information altitude, mostly just seeing Big Data landscape features: the information equivalent of a mountain range, an ocean or a city. However, the data should be comprehensive enough for management to spot any digital puffs of smoke coming from fires that need to be put out. A Balanced Scorecard with performance indicators, trends and information on the data’s strategic importance would be the best tool here. If any one indicator is performing below expectations, management needs to dive into it to understand what’s causing the problem.

Once a problem is identified, data presenters and their software must be nimble enough to zoom in to provide a closer look for the appropriate managers. When a problem is spotted, management will want to drop down to 20,000 feet. At this point, the data presenters need to narrow the scope of information and begin providing some operational tools, such as flow charts, to gain an understanding of what’s happening.

Next, data presenters must be able to descend to 10,000 feet and provide diagnostic tools so managers can become more directional in their behaviour. This isn’t meant to be eye candy. It should be a detailed, data-intensive dashboard that provides comprehensive information designed to help managers at a tactical level.

Finally, at ground level, organisations need analytic tools so management can become prescriptive. These are the spreadsheets and grids that display all forms of detailed data used to assess how the organisation is performing individual tasks.

There’s a parallel between the levels of the organisation. In a good organisation, the leadership team should be able to pass this information to the next level down and so forth.

Brett Knowles ( is the executive partner of pm2, a performance management consulting company based in Canada. He has assisted more than 3,000 organisations around the world.