Three ways businesses can manage risks and rewards of data

Leaders often like to say their business is data-led. Of course, data is a hugely powerful resource for businesses, but if it is not managed and interpreted cleverly, it can wreak havoc within an organisation.

Why? Well, the data might turn out to be wrong (for example, as it did at the Post Office). The data might be incomplete, such as when a retailer focuses only on customer satisfaction, thereby ignoring people who don’t shop with them at all. And the data will be historic – there is no data of the future. For many years, data showed pandemics didn’t cause that much trouble in the UK.

So data brings risk as well as reward. How can we manage the risk, whilst enjoying the reward? Here are three ways.

Think hard about targets

Focus on the outcomes you seek. For example, be careful about targeting brand penetration – you can’t put penetration in the bank. Then make sure you target the long as well as short term. For example, a sales target for this year might lead to you paying too much for new distribution, hurting long-term profit.

Set targets using data from a variety of sources, so if one measure misses a market dynamic, you pick it up with another. A classic example is that many companies underestimated discounters for a long time because they weren’t included in the scanning measures. Targets drive behaviours, so think hard.

Celebrate inconsistency in your measures

Use it to ferret for threats and opportunities. If two measures seem to disagree, don’t pick the one that makes you feel more comfortable. Focus on the ugly one.

If your data says you’re growing fastest in the retail market but then a competitor reports faster sales growth, don’t assume they are cooking the books. Maybe their growth is coming from foodservice and you’re not tracking that? Maybe that’s an opportunity?

Continually check in with human beings to test your data

This means talking to your consumers. I don’t mean conducting a survey, I mean leaders personally talking to consumers. It also means talking to consumer-facing colleagues (such as store managers for a retailer).

If your call centre’s customer satisfaction score looks good but your friends tell you about bad experiences and your colleagues highlight problems, it’s more likely your measure is corrupt than that you keep getting a bum steer from humans.

A common corruption is frontline staff, under pressure to “deliver the satisfaction numbers”, briefing customers to “please say you would recommend us, when you fill out this survey”.

Of course data is powerful, but it can cause big problems. To manage it, obsess over the right targets, celebrate inconsistencies, and check in continually with human beings.