Number crunching

In the race to implement the tools to capitalise on the rapid expansion of available data, we sometimes forget why we are bothering. It’s not because we like looking through thousands of lines of data. Or because we want people to complete fields with data that they think is irrelevant. It’s because we want to understand the world we live in so that we, the humans, make better decisions.

Procurement technology and data is a familiar subject to regular readers of this blog. I wish to return to it now to explain how to approach. With the help of the IBM business analytics blog I’ll describe each stage and show how each step builds on the output of the previous stage.

Before going further, we should take a moment to recognise the place of humble spreadsheets in the history of business software. Spreadsheets are the original killer app because they are easy to use and ubiquitous. They are, however, disconnected, siloed, and ungovernable which means they are not reliable, and they cannot scale effectively. If procurement really wants to make the most of big data then an enterprise-grade approach is requires. Data scientist should be involved at each stage of the process and use dedicated, connected tools that help collect, prepare, analyse, adjust and present data to decision makers.

Planning analytics What is our plan?

It all starts with a plan. Whether it’s the corporate plan, a category plan or next year’s budget. It requires an understanding of past performance, identification of deviations from the norm (plan vs. actual), evaluation of possible scenarios, prediction of likely outcomes, and assessment of risks and constraints.

Descriptive analytics What happened?

Now that we’ve got a plan, we want to know what’s happened. Descriptive analytics use two or more historical data points to illustrate a trend, for example, the spend on IT contractors in 2017, 2018 and 2019.

Diagnostic analytics Why did it happen?

Diagnostic analysis takes the descriptive analysis, adds third party data and provides one or more explanations for the changes. Continuing with the example of IT contractors, agencies that engage contractors in the public sector – rather than the personal service companies that employ the contractor – became responsible for applying IR35, the United Kingdom’s anti-avoidance tax legislation designed to tax disguised employment at a rate similar to employment, from April 2017. Some IT contractors in the public sector found their take home pay cut significantly while they were expected to do the same work. Many terminated their contracts in March 2017 or declined renewals and looked for work in the private sector companies that were unaffected by IR35 at the time. While demand remained the same, supply increased leading to a decrease in day rates in the private sector.

Predictive analytics What will happen next?

Predictive analysis takes diagnostic analysis and applies it to the future. IR35 is going to be introduced in the private sector from April 2020. We are likely to see a couple of different outcomes: some IT contractors will look for permanent roles while others will try to negotiate higher day rates to offset the greater tax burden. This will put private companies under pressure but supply and demand will stay the same so there will be no economic reason to increase day rates.

Prescriptive analytics What should be done about it?

Prescriptive analysis takes all the previous analysis and suggests actions to achieve certain outcomes. One of these outcomes might be to reduce the cost of providing the current level of IT support. One suggestion would be to offer IT contractors who have been with the organisation for a long time the option of a permanent role at a lower overall cost (pay and benefits is less than the day rate). It might be to outsource more of the IT support and additional tax liability to a third party. Finally, it might be to seek IT support from outside UK thereby avoiding IR35 altogether.

Conclusion

The number of organisations using each stage – descriptive, diagnostic, predictive and prescriptive – decreases sharply. As the value of later stages becomes more apparent and easier to achieve, we will see more organisations using diagnostic, predictive and prescriptive analytics. Afterall, as the British philosopher Carveth Read said, “It is better to be vaguely right than exactly wrong.”

If you enjoyed this article, please consider sharing it!
Icon Icon Icon