Archive: April, 2018

3. Data differences

This is the third in a series of five blogs taking an in-depth look at procurement technology and data. I started by considering the range of procurement technology that is available. In this blog I compare data generated by procurement systems and social media.

It seems not a week goes by without another revelation about the use of data. The latest story to hit the headlines is Facebook’s harvesting of personal data. If some of the research is to be believed, “it just takes 150 Facebook likes for psychometrics software such as Cambridge Analytica to know your needs, fears and hopes better than your parents do, and just over 300 likes for such software to know you better than you know yourself.”

And yet everyone in procurement seems to be frustrated by the lack of the good quality spend data and the way it’s presented. This frustration seems to be felt most acutely in organisations that have invested heavily in enterprise resource planning (ERP) software.

So why does personal data appear to be having a much greater impact than spend data? Are Google and Facebook better at handling big data than SAP and Oracle? Or did large organisations realise the value of data a long time ago and consumers and governments are playing catch up with regulations like GDPR?

I think there are a number of reasons: the availability of data, the quality standards applied and the way in which it is used.

Personal data is gathered from a variety of sources: search engines like Google, websites like Amazon, and location based services like Garmin. As more devices like fridges are connected to the internet, even more data will become available. Spend data, however, is only created when transactions are made in ERP systems. These rely on accurate master data on users, suppliers and contracts and high levels of compliance to processes and controls.

When researching this blog I was surprised to learn that “most marketing data is between 10% and 20% accurate.”. Without doubt, Facebook and Google’s data is a lot more accurate but it hard to get concrete evidence, for example, there are claims that Facebook’s gender data is 99% accurate. This figure comes with a caveat that there is “a bot problem”. So is it 99% accurate or not? From a personal perspective, I’m occasionally surprised at the accuracy of online ads but more often bemused at how they ended up in my feed.

I cannot imagine sitting in front of a CFO trying to explain that only a fifth of the spend data presented was correct. Furthermore, I’d expect suppliers responding to a tender to build in a significant risk premium to their pricing. I’m not saying that spend data is perfect. Even investing a lot of time improving the quality of spend data, I often apply Pareto’s principle, that is, 80% of the effect comes from 20% or the cause. But spend data has to give a reasonably accurate picture because budgets and savings targets are often based on it.

Big data has changed much in the world of advertising but John Wanamaker (1838-1922) quote still seems holds some truth: “Half the money I spend on advertising is wasted; the trouble is I don’t know which half”. Big data has also changed much in the world of procurement but for most organisations big improvements are still required. It seems we all face the same challenge but have different ways to tackle it.