All forecasts are wrong
By nature, all forecasts are wrong; the key is understanding and minimising that error because significant forecast error impacts productivity and costs of operation.
But tackling this can seem like a mountain to climb, especially when you’re stuck in the monthly trading cycle. Operations face a crisis and dealing with day-to-day issues takes precedence. Just look how hard it is to find enough people. Logistics UK reports that, as of 2022, some warehouse operations vacancy rates are over 20% and that chronic worker shortages result in offers of 30%+ wage increases.
Modern supply chains built for efficiency, not resilience
It’s also true that supply chains are more fragile than they used to be; since the last recession in 2008, supply chains, logistics and operations have been built to be efficient, streamlined and to minimise cost – what this means in practice is that they may be cheaper, but they’re not resilient. One thing, as simple as a spike in volumes, can cause a fracture. Amazon have reset the industry benchmarks for same day/next day delivery, exerting more pressure on supply chains and their abilities to meet the customer expectation.
One way we’ve helped logistics and operations teams is to develop forecasts that are fit for their purpose. That speak their language. And here’s how.
In the battle of top-down versus bottom-up forecasting, there can be a meeting in the middle. A forecast that actually helps to better plan labour, space and resource rather than nod to high level corporate forecasts.
At first, this might seem to be at odds with the purist view that the whole business should work from ‘one version of the truth’. But so long as the different forecasts are being used to inform relatively independent decisions and it is understood that they may not be directly comparable, this approach can add value. The key is finding the linkages – how does the average line value change over time, do the lines per order fluctuate, do the units per pallet flex with seasonality, what impacts do these factors have on output per hour?
Working with a national supermarket chain, we did the following:
Rather than attempt to take a forecast generated at SKU / store / week and aggregate to meet the needs of the warehouse operations, we produced a seasonal forecast (using Holt-Winters) at product category / site.
The reduced volatility of this bespoke forecast meant we could more accurately predict demand and better manage resources and costs. This better planning and more realistic forecast also meant we could better cope with fluctuations and be more agile on the ground. In a word, we were more resilient, but without adding extra labour cost in as a buffer.
Other things that can help to ease the squeeze on space and costs
Ongoing difficulties in finding enough people means companies are looking to automation as a silver bullet to solve their forecasting woes by building in redundant capacity. But full automation comes at a high capital price, so mechanising or partially adopting automation in particular processes can offer a resilient, fixed level of throughput capacity which is supplemented by manual operations.
Fit-for-purpose forecasting and monitoring can be invaluable in the scoping of new technologies. If you are about to invest significant CAPEX in automation, you want to have confidence that the business projections you’re basing that decision on are sound.
2. Incentivising buyers and trading teams to manage excess stock
It’s surprising how much unsold, excess stock is sitting in warehouses, sometimes for years, perhaps even decades. All this stock is costing money while it’s sat there – space, energy, and people (just think about the annual stock-take).
Perhaps in a way it’s not surprising – no one is incentivised to get rid of it! Buyers are often measured on gross margin, trading teams on total sales by category. Profit per SKU is calculated on cost to buy and how many sold – it’s not often things such as space, energy and labour are factored into these equations.
To resolve this, companies should initiate product life-cycle processes. Capturing trigger points of when SKU’s start the decline phase, ensuring the mechanisms and reporting are in place to discontinue the items to prevent replenishment, and start the promotional activity to deplete any inventory.
Financial reporting to capture the depreciation and accrue a provision is a great way to assess the quality of NPI investments over time and incentivise buyers to make better decisions in the first instance.
The final stage in the lifecycle should be more aggressive – financially writing off the SKU and selecting the best route for exit – broker, charity, or the cardinal sin, disposal!
As companies face an unprecedented dark winter, shifts in demand and buying patterns will call for an agile and resilient response. Purpose-led forecasting can help businesses meet and overcome challenges they will face in the future.