Supermarkets are returning to the kind of forecasting activity which has previously led farmers to plough millions of tonnes of produce back into the ground, despite the advent of high-tech forecasting tools.
A survey by the Groceries Code Adjudicator found poor supermarket forecasting behaviour had become “more entrenched” despite the growth of new supplier portals.
It also found some retailers were charging large sums for suppliers to have access to their forecasting data, some of which was previously free.
‘Unclear, unreliable’ data
The GCA survey found suppliers were being confronted with “unclear, unreliable and unusable” data from supermarkets against an increasingly “volatile” backdrop.
Detailed interviews with 28 direct suppliers, carried out following the GCA’s survey in June, showed what YouGov researchers called an “emerging pattern” of behaviour.
It has raised fears that supermarkets are again engaged in the kind of activity which in 2017 saw them slammed in a report by previous Adjudicator Christine Tacon for “systematic failures” which were causing huge levels of food waste.
This year’s survey found 17% of suppliers said they had incurred significant costs because of inaccurate forecasting from retailers, down from 18% in 2024.
Despite the slight fall, YouGov research manager Lois Harmer said interviews had uncovered a growing pattern of “erratic” behaviour.
Forecasts were often unreliable or didn’t match what was ordered, added Harmer.
The survey found retailers were often unwilling to accept responsibility for forecasting issues, with the onus on suppliers to deal with unsold stock or to supply stock at short notice.
“One supplier summed it up saying ‘they don’t order for two weeks, then suddenly they want 10,000 units’,” she said.
Another anonymous supplier said: “Forecasts can be quite volatile, especially on high-performing lines and at promo periods. They have continually been under-ordering, so then our availability to promise is too low.”
A third supplier source said there was “zero accountability” from retailers to rectify problems.
Harmer said there was too little flexibility, with suppliers of fresh produce and imported goods particularly exposed because their margins were so tight.
She added: “Some retailers now charge suppliers for access to sales data, but without access to data it’s impossible to forecast accurately.”
Forecasting best practice
In March 2016, the GCA published a statement of best practice on retailer forecasting. Further guidance was published after the 2017 report by Tacon, in which she uncovered “horror stories” of crops being ploughed back in to the fields because of last-minute changes to orders.
Adjudicator Mark White told last week’s conference he was making issues relating to forecasting a “priority”.
He said many retailers were currently updating their forecasting capacity using the latest AI.
“As they make these upgrades which should benefit both suppliers and retailers, the success of the transformation will largely rely on how retailers explain the new system to suppliers,” he said.
Retail Mind founder and GSCOP expert Ged Futter said he was calling on the Adjudicator to obtain a signed declaration from supermarkets that they would uphold best practice on forecasting.
“We’ve had previous best practice statements on this but the time has come for further action. We’re seeing a huge amount of inconsistency.
“Some retailers work collaboratively and some don’t. Some just rely on their own systems and those systems generally work a couple of weeks in advance. That’s not enough for suppliers.”
Futter added: “One of the things I’ve heard is that retailers are asking suppliers to hold onto more stock because they want to reduce the stockholding in their own depots.
“At every opportunity retailers are pushing the risk back to suppliers and if suppliers want store-level data they have to buy it.
“Is it any wonder we’re seeing gaps on shelves?”
David Sables, CEO of Sentinel Management Consultants, said he believed supermarkets looking to open up streams for more revenue under retail media may be contributing to the problem.
“This is an absolute nightmare for categories that are seasonal or short-lived.
“The fact of the matter is you need to have a joint approach to forecasting and that isn’t taking place. The reality is you could count on one hand the amount of genuine joint-forecasting that goes on.
“I think the AI that Mark White talks about will make it better in the medium term, but it’s not going to come to the rescue quickly enough to help those who are going through this forecasting nightmare right now.”
No comments yet