Article
Written by Barry Delhez
Posted on 28/01/2018

Predictive analytics: making the unpredictable predictable

271 reads

Predictive analytics is the crystal ball of the business world. It’s a technique that enables you to predict what the likelihood is of a customer becoming insolvent, as perhaps first suspected. Alternatively, it helps you decide whether to do business with  a promising start-up. A pre-warned credit manager is worth its weight in gold.

Predictive analytics begins with gathering data. This data may come from your own company. Over the course of the years, companies generate masses of data about their own customers: their loyalty, number of contracts, turnover per customer, visits to customers, etc. This data, held in-house, provides interesting information, although it does have one major disadvantage. It is often very subjectivewhereby it only tells you some of the things about your own customers. So, the best course to take is to supplement your in-house data with information from the external sources. Gather data about your competitor’s latest product, pick up comments from the social media, check to see how the market is developing and so on. The more data you can gather, the more accurate your prediction will be.

Links and correlations

Do you have enough data on hand? If so, go looking for patterns or correlations. This is precisely the essence of predictive analytics: that certain variables actually strengthen one another. For example, you feel intuitively that there is probably a link between the orders placed by a customer and the number of visits they receive from their account manager. Companies that see a salesperson more often place more orders. Admittedly this is an obvious link, however there are definitely more patterns and relationships you could be noticing which may be slipping under the radar. Predictive analytics enables you to tease these links out.

“Compare your average customer profile with your prospects to see whether you should work with them or not.”

That way you can predict accurately how likely certain new business prospects will really become your customers. What actually happens is that your customers tend to be very similar. So compare your ‘average’ customer profile with your prospect. The more their features correspond, the greater the chance of you working with your lead.

The greater the frequency, the higher the accuracy

An interesting fact is that the more often you use predictive analytics, the more reliable the results become. You can also check to see whether the predictions made actually turn out to be accurate. You can then add those results to your existing dataset. This makes the system increasingly intelligent and the predictions more and more accurate.

“Predictive analytics helps you anticipate future events.”

Gathering data and exposing insights is both enjoyable and surprising, but is doesn’t help your company gain any ground. You need to act promptly if you are to avoid certain predictions coming true – or simply exclude them altogether. Predictive analytics helps you to anticipate future eventsfor a better outcome had you not taken any action.

A good example of predicative analytics can be seen in practice at Walmart. They discovered that when bad weather was approaching, shoppers also bought gingerbread as well as spare batteries and pocket flashlights. This information enabled Walmart to generate more sales. As a result, when the weather forecast is poor, Walmart fills its shelves with additional gingerbread...

Augur Score 

Graydon’s Augur Score predicts the likelihood of future business failure within a 12 month period.. This information is crucial for Financeand a company’s bottom line cash flow forecasting. Accoodingly, the exposure to business risk is mitigated.

“Which company will no longer exist in 12 months’ time?”

Previously the assumption was that around 5% of companies would cease to exist each year. Out of turnover of 1 million euro, that meant a risk of 50,000 euro. The Augur Score defines what proportion of your turnover is threatened on a debtor level. The score may show that debtors with a higher chance of going out of business have a higher average turnover. If that is the case, you will need to take account of a possible loss of 80,000 euro instead of 50,000 euro. This and other insights are also of interest for sales and marketing, as well as for credit management. So, by no longer spending your budget on companies that are on the brink of business failure, you will increase your returns.