Predictive and Prescriptive in Consumer Packaged Goods (CPG) Across M&A and Divestiture Scenarios

Predictive and prescriptive analytics starts with reliable data and comprehensive and accessible lineage and history.

Predictive and prescriptive analytics starts with reliable data and comprehensive and accessible lineage and history.

What is one of the biggest hobgoblins of CPG when it comes to M&A and Divestitures? Data history. 

Here’s what we mean. Let’s say Company A measures the profitability of the products it produces by what plant it is produced in.

Let’s say Company B measures the profitability of the products it produces grouped into Brands at the average it costs to produce across plants or sub-contracted producers of the product.

Now let’s say Company B acquires Company A. What happens? They integrate those products and the way they look at the Brands going forward in the way they have always looked at things—once integrated, on a go-forward basis. What about the history of Company A and how its products performed over time? Often lost, or watered down to the basics or just revenue, units, and some aggregated Trade and COGS metrics. It may then be years before Company B builds up enough history to make well-informed decisions on the newly integrated company’s products—and then to be able to reliably predict or inform decision support on next actions to take. Even if Company A and Company B had measured and analyzed profitability similarly, there’s a good chance in an acquisition/integration scenario, key and useful history would still be lost. 

This happens quite frequently in CPG where brands may be divested from large companies and purchased by other large companies or combined together or broken up by venture capitalists. Or when a small startup CPG company gets acquired by a larger player. History and valuable data is lost all the time as the acquiring company may decide it is too difficult, too costly, or too time consuming to somehow recast the historical data to fit how it looks at measuring the acquisition target’s goods. Sometimes, it is even considered a “risk” to an integration project. Or it is prematurely concluded that what is needed just does not exist at the acquired company.

Without an adequate base of apples-to-apples comparably cast history, it’s hard to make decisions and have any idea how accurate a prediction can be made—or what decisions can be made—other than by basic inference.

One thing we do all the time at Hybrid Intelligence is provide a bridge between systems and companies in mergers, acquisitions, and divestitures. It can be done. Moreover, often there are needs to be technical and automated mechanisms put in place to continually recast for a period of time based on new information, data, or observations noted as the business does its analysis. It can often not be done very effectively by a traditional IT or business project team which may lack the specialization or resources required for this delicate, sometimes complex, and yet very important task.  Historical data conversions, bridging, and recasting for comparability may just not be deemed that important—an afterthought, an annoyance—“Oh yeah, we have to handle the history, too.” That conclusion often comes from systems-type people and not business folks who need that data for business analysis. We posit that this historical bridging between systems and companies is one of the most valuable and important things that can be done to improve data analysis, analytics, P&L planning and budgeting—and particularly useful in predictive and prescriptive needs as brands may be acquired and integrated at any CPG company.