The accuracy of the demand forecast has a huge impact on the supply chain – suppliers, warehouses, stores, logistics and transport…
Fundamentally it all boils down to robust maths, where statistical algorithms are used to calculate the most likely future demand at whatever level of aggregation is appropriate – from as low as sku store day through to various aggregations across product location and time hierarchies.
The only certainty in forecasting is that you will never be 100% accurate!
So, the basis behind demand forecasting is simple really – it’s just a bunch of various statistical models, forecast algorithms and calculations. Massive data volumes being pumped though number-crunching engines. In the end though, no matter how hard you try, you can only be certain of one thing… No matter how good you are, or how fantastic your systems are, the demand forecast is going to be wrong.
So, why bother with forecasting at all then?
It’s not the most exciting place to be – to know that the only certainty is to be wrong – so, what’s the point of forecasting at all then we hear you ask?!
I’ve learnt that for most retailers it is understanding the measure of the inaccuracy is actually what really matters. If you are able to provision for the risk of out of stock or over stock as a result of the inaccuracy, then you have the key to unlocking your supply chain opportunity through a quality demand forecasting process.
When the Skilled Demand Planner can really add value…
Once the products have been signed off as a result of the range and space planning processes your demand planner really comes into their own… When provided with information such as:
- Current market behaviour and range forecasts / range direction
- Past and current range performance / range trends
- Historical analysis of the impact of any demand influencing factors
then a skilled forecaster can estimate, top down, the forecast flow (demand-pull) of a product using additional inputs such as:
- Estimated rate of sale for the range
- Estimated product mix / share of total range sales
- Past performance / demand patterns of similar items
- Seasonality curves
and can achieve a fairly high degree of accuracy – certainly adequate for starting cost price negotiations based on volume commitments for example, not to mention having the necessary supply chain flow data for populating distribution capacity plans and ensuring adequate capacity is secured to cope with key seasonal peaks.
Accurate Sku-Store-Week Demand Forecast – A Panacea or a Possibility?
So far it’s all been a relatively simple task for your demand forecaster (whatever they’ll have you believe!) The real complexity arises as the product is received into your supply chain and requires further distribution down to your stores.
Questions around which stores will perform best for which products are asked, and at this level of disaggregation the data tends not to be as statistically valid for most forecast algorithms as it would have been at the higher levels of aggregation.
To achieve an accurate sku store level forecast, then to cut that further down into weekly, even daily, time buckets, would be the panacea for your demand planner. It would enable numerous benefits including:
- The ability to sense check the level of display stock set into the system – enabling debate over assortment planning decisions and highlighting any issues with store assortments, by exception (e.g. where a store’s minimum display stock exceed X weeks cover based on the detailed sku-store demand forecast)
- The ability to run a completely bottom up, pull based replenishment plan or to create an ‘order up to level’ driven supply chain plan.
Don’t let demand forecasting be the ‘Poisoned Chalice’ of your business processes…
My team at Retail Acumen recognise that demand forecasting is one of the most data hungry processes in retailing, where data accuracy and data quality have massive repercussions, and where simple errors in processes such as a store physical count or an inaccurate order lead time setting can result in a store residual stock problem which can ultimately erode the net delivered margin of a product significantly in just one inaccurate user’s key-stroke!
However, it’s not all doom and gloom, demand forecasting need not be the poisoned chalice of your business processes…
The blatant plug 
Retail Acumen can help you to develop clear processes, to cleanse and improve the quality of your data, to up-skill and educate your teams on the concepts and the importance of demand forecasting to the business; all enabling you to realise significant benefit and increased profitability from your improved forecasting capabilities.
To find out more, please contact us. We forecast it’ll be well worth it!
Two initial key factors in the ‘early life’ of a product often act as the seeds of later chaos: poor (often over optimistic) estimation of rate of sale and lack of coordination between display merchandising and demand planning. Best sellers are less of a problem but often get most focus to avoid the dreaded threat of stock-out, visible to everyone. It is in the slowest selling third of a category or range that the problems are more often left to fester with variable stock levels, often excess to a desirable, net profitable level.
In the planning stages of a new range or category revision it is often neglected to address this ‘duly diligent’ planning at an intermediate level, on a human scale of judgment between the awesome detail and mighty processing power of the supply system and the majestic vision and sweeping generalisation of the range review.
In my experience large boosts in productivity, profit and large stock savings are discovered in this territory.
Sales and Operations Planning is a good initiative to do to improve the whole planning process.
Pingback: Supply Chain – Planning, Managing and Communicating | Retail Acumen: Delivering Excellence in Retail
Hi Clare,
Thanks, as always, for an informative and thought provoking blog.
There’s a couple of dimensions I’d like to add, which I think are important considerations;
Communication – Once your Demand Plan is formed – Share it with your Suppliers. The information you’ve gathered will be invaluable to your suppliers in trying to forecast the demand you will place on them and will help avoid the risk to continuity of supply and equally as important for the supplier, the risk of overstock. Both drive unnecessary cost into the Supply Chain which drives unneccessary cost back to the retailer. This is particularly important when planning promotions. Recognising the need for complete confidentiality, your supplier should already have signed a Non-Disclosure Agreement or Confidentiality Agreement, work in partnership with your supplier, share your impact forecast and jointly reap the benefits. My bad experience? Sitting at home on Thursday night watching advertising from major clients advising of promotional weekends which drive unprescedented sales knowing I could support the additional demand!
The second aspect? Make sure your Demand Planners understand the customer purchase. This is particularly important when considering ‘project sales’, such as paint, tiles, etc where multiple units to undertake the purchase. Demand Planners can add enormous value to the Retailer but can, if not inducted effectively, rely on pure Mathematics that omit the customer purchase profile and the replenishment cycle. This may seem basic but again I’ve first hand experience of a retailing major client advising us of the replanned store profile which had been profiled based on the smoother demand profile. In this example the Demand Planner had smoothed the demand showing by example that a product sold an average of 2 units per week, therefore they would fill to a maximum of 4. What the planner did not know was that the average sale was 5 because 5 was the quantity that a customer required to undertake an average project. Once again, communicate, a good supplier will know their product and the customers buying habits, use them to help educate the planning team and increase availability and customer satisfaction.
Warren Dow, Managing Director, Partners In Supply. Supply Chain and Supply Chain Risk Management Solutions
Thank you for your thoughts Warren and I agree with what you say – in fact my next 2 articles on supply chain do explore beyond the forecast itself into the planning, communication and collaboration aspects!