Supply chains crisscross the globe and bring together goods, technology, people, innovation, training, and investment to manufacture products that most people couldn’t access, or afford to buy, a few decades ago. Supply chains form the backbone of the global economy. And when times are good and a company’s supply chains are flowing predictably, a lot of management focus is on market growth and cost optimisation.
Things don’t always go smoothly and disruptions in our supply lines are common. With mature supply chains, the challenges faced along the nodes of it can often be resolved without customers being aware that their orders required circumventing obstacles in order to make the delivery deadlines.
But what happens when there is a number of events that your business has never encountered before? And what happens to your planning operations when the world is faced with a Black Swan event?
Since COVID I have heard people lament “What is the point of forecasting, we can’t possibly get it right”, I have also heard the same said of the annual budget process.
Imagine you’re driving a car at night on a dark country road, and the task of getting to your destination on time is analogous to keeping your customers satisfied.
If you stop or crash you would force your customers towards your competitors.
Would you stop looking out the windscreen because it’s dark and you can’t see clearly?
Of course not. Most of your attention would be looking ahead, using any clues to guide you regarding the state of the road, the direction of the road, and any pending dangers.
The suggestion that we should not bother to forecast is like saying we should shut our eyes while driving, because it is dark outside.
This is the time to up the ante with your forecasting frequency and ensure you stay ahead of the game.
How can you best achieve this? (Hint – there won’t be much looking in your rear view mirror.)
1. Lift your line of site (Keep your eyes on the horizon)
In times of uncertainty we often see management focusing on the now: Focusing on last month’s sales, month to date sales, and today’s new sales.
Tracking these metrics is essential for situational awareness (like looking at your map or dashboard to confirm where you are and whether you are going fast enough to get there in time.)
Yet using these metrics alone is akin to driving by looking in the rear vision mirror. Motivated by our self-preservation we know instinctively to lift our eyes and scan the edge of the visible horizon.
To preserve our businesses we must do the same and forecast beyond our order horizon.
The expectation of this forecast is not to be perfect 12 months out from now. Our forecast is to build a working estimate of the road ahead (our future demand) in a context where we can capture all market intelligence and known changes and apply any changes as we become aware of them.
This allows us to navigate impacts in demand as early as possible.
2. Involve and challenge your customers (don’t ASSume)
Your biggest customers are a great source of prediction so it’s important to involve them in your forecasting operations. No matter how good your forecast processes are your customers promotions will be planned in confidence and often only their marketing team will know when and what the impact will be. It is folly to go it alone.
Ask your customers to submit forecasts.
Often I hear businesses say, they won’t give us a forecast. If they won’t, then provide them with your forecast of their requirements and ask them to ratify it.
Arrange to review your customer’s forecasts with them. Just because they gave you a forecast doesn’t mean it has been well considered and error free.
If your largest customer gave you 6 months of forecast would you assume production stops in the 7th month? Maybe it will! If it contained discontinued SKUs are they relaunching or should the forecast be corrected? How will you know unless you ask? Invite them to participate in your demand review process and review their forecast with them.
3. Use statistics to frame your forecasts (but don’t over value them)
I am fond of reminding students: Statistics are the next best thing to understanding your customer’s needs.
With the availability of big data, machine learning and assorted analytics there has been an unprecedented level of tech companies and their marketing departments generating hype – advising us of how AI is changing the world. One of the tangible areas that this can gain traction is the forecasting of large sample data and particularly demand sensing.
These tools are expensive and far from fallible, but they can offer tangible improvements in forecast accuracy particularly if you are B2C (business to consumer).
An example where you would use statistics, and even big data and machine learning algorithms, to determine what products and how much to replenish, is at a supermarket.
Consider. A supermarket store can’t ask all 80,000 customers for their next 12 month’s grocery forecasts. In this case where the number of customers is too vast to have intimate and frequent dialogue, statistics provide a method by which we can use recent sales to predict future sales.
However If you relied on toilet paper sales statistics from March this year you would have seen demand sensing algorithms suggesting that the whole of Australia had attended a month long vindaloo eating convention. Further long range forecasts will show massive demand spikes in March of future years.
Unless you can apply customer domain knowledge (customer context and common sense) to the statistical information, you would risk replacing all items in the supermarket with just toilet paper.
In that example, the customer use of toilet paper had not increased, only the short term stockpiling, so the anticipated response would be that toilet paper sales would reduce in subsequent months as consumers let their emergency stockpiles return to normal levels.
If your customers are businesses you can ask your top 10 largest customers to provide forecasts so you can ensure you have reserved capacity and material for their orders.
When you have conversations about any unusual trends they are seeing in their demand patterns you will get closer to the true source of their concerns and get lead indicators of changes that they are driving in the market such as new innovations or product launches. Your statistics can only respond after the sales have happened.
Statistics are not always correct, and machine learning is not magical (no matter how often marketing repeat the word AI). They are impressive tools that can help show trends and quantify impacts.
Their drawback is when businesses adopt these tools and then ignore their major customers in the forecasting process. The customer intimacy is developed by doing demand reviews with your customers. This should be preserved at all costs.
4. Add market intelligence (you know more than you realise)
Everyone has an opinion, a piece of the puzzle, an eye for a storm. So ask them! You may be stunned to learn the deckhand knows more about the weather patterns than the Captain (his dad is a meteorologist.)
The supply chain manager who has been locked in his office may know less than the truck driver because the truckie overheard the customer at the receiving dock talking about their plans to sell out.
We have worked with a number of businesses during COVID which had staff who were adamant that they couldn’t forecast in the current climate. Yet with just a little bit of coaching, we were able to breakdown their portfolio of products and customers that were:
1) Doing a bit better with COVID,
2) Doing a bit worse with COVID
3) Not trading with COVID
Once we had agreed that general categorisation, which there was no argument with, the sales team were able to forecast for each of their customers with some confidence. People feel more valued, thus work harder, when asked and are involved in the process so their output may also increase (win win!).
5. Consensus rules (sure beats confrontation!)
Collaboration is invaluable to get the best market intel and information. Speak to the concerned stakeholders. Share observations – this can be more vital than the forecast itself. How well considered your forecasts are depends on how many people use all the available information well. It is very liberating for businesses to work together – momentum is comforting and who doesn’t like some reassurance they are on the right tack? You can’t stop dead in the water because you’re not sure where you’re going – you have to use the best navigational tools available to plot the best course to ensure you win the race.
Author: Tim Gray
To hear how entrepreneurs around the world overcome their challenges, search your favourite podcast platform for ‘CallumConnects‘ to hear a 5-minute daily breakdown.
Callum Laing is an entrepreneur and investor based in Singapore. He has previously started, built, and sold half a dozen businesses and is now a Partner at Unity-Group Private Equity and Co-Founder and CEO of MBH Corporation PLC. He is the author of three best-selling books ‘Progressive Partnerships’, ‘Agglomerate’, and ‘Entrepreneurial Investing’.