Twenty basis points in ten years

A flat category changes what you are allowed to sell

The opening frame was deliberately unflattering. Food and beverage is enormous, a value chain carrying over 15 trillion dollars of revenue in 2015 on the third party estimate used, close to a quarter of world economic activity. It is also, on the European consumer spending data, essentially flat as a share of wallet. Two tenths of a percentage point in a decade.

A supplier selling into a flat category cannot lead with growth. Whatever it sells has to release money that is already in the system, which means finding where the manufacturer is losing it. Two forces do that. Retailers press on the top line, suppliers press on the bottom line, and the manufacturer margin sits between them.

So the work was structured backwards from margin rather than forwards from technology. Four drivers, eight trends, and one question against each: which of the five business functions does this actually hit.

The function everyone pitches at scored lowest

Each of the eight trends was scored one to four against product development, production and packaging, sourcing and supply chain, sales and marketing, and consumer engagement. The scores are a judgement made by the team, not a measurement, and they were made before any solution was drawn.

The fours clustered on sourcing and supply chain, driven by local sourcing, food safety and recall prevention, and reformulation for indulgence. Consumer engagement picked up the convenience and direct channel trends. Product development, which is where a technology pitch instinctively goes, was scored once in the entire matrix, at the lowest level on the scale.

That is an uncomfortable result for a supplier with a product development story, and it was the most useful thing the matrix produced.

Test the themes against what customers already published

Four renewal themes came out of the scoring: analytics led decision making, voice of consumer, operational performance, and the supply chain balancing act. Rather than argue for them, each was tested against the published strategic goals of five Nordic producers, taken from their own annual reports, with every goal tagged to a theme.

The benchmark work underneath was blunt. One meat and poultry processor was running research and development at roughly a third of its closest peer while its published strategy promised an ambitious product pipeline, and its return on assets was negative in the most recent year reviewed. A dairy business showed three consecutive years of declining sales growth on an operating margin well under its international comparator. Both readings are stated in the deck as hedges, might signal and might indicate, and they stay hedges here.

None of those companies is named on this page. The benchmark is published accounts restated by the team, and the conclusions drawn from it are unflattering, so naming them would be somebody else reputation spent on our marketing.

Where the money actually leaks

The cost case for traceability and waste work turned out to be the strongest, and it is built entirely on third party industry figures. Product recalls had risen fourfold over five years with an average direct cost of ten million dollars. Food fraud was costed at ten to fifteen billion dollars a year to the industry, described as potentially preventable rather than recoverable. Raw materials run at 60 to 80 per cent of material cost and waste at 5 to 8 per cent.

Set against a flat category, a few points of waste is a larger prize than a few points of volume, and it is a prize the manufacturer keeps rather than shares with the retailer.

The counter example on portfolio complexity was the one that stuck. Beer market stock keeping units had grown from 300 to over a thousand chasing differentiation, while one category in one European market grew revenue 17 per cent while cutting its range by 42 per cent. More variants is not the same as more sales.

Three concepts, ordered by what the industry can absorb

The output was three solution concepts placed on a horizon map rather than a single recommendation. A consumer interaction engine first, because consumer signal is dispersed and reading it in one place reduces failed launches. A shared business to business data hub second, moving the chain from making to stock towards producing to demand, which attacks stock outs, shelf expiry waste and inventory carried at every step. A distributed traceability ledger third.

The third one was the one everybody wanted to discuss, and it was placed last for a reason the deck states plainly. It needs a critical mass of participants before it works at all, awareness in the sector was low, and every transaction on it can be traced back to an identity, which is a privacy problem before it is a technology problem.

What the client was left holding, and what it did not get

A scored map of eight industry trends against five business functions, so a sales conversation can start where the customer is losing money rather than where the product is strongest. A benchmark of five producers against their own published goals. And three concepts with the stated challenges attached to each, including the ones that would kill them.

This was desk research, not fieldwork. No interviews were run, and every external figure in it belongs to a published report or another company annual accounts. Nothing was built, nothing was piloted, and there is no measured result in this engagement at all. What it produced was a defensible reason to knock on a particular door with a particular thing to say.