Path-finding in a time of crisis
Crises are overwhelming. A so-called “perfect storm” crisis can easily finish a business. Hindsight then tells us to look back and think what we might have done, what we should have done, if only. Exactly because of this, we sometimes dream of being able to see into the future, so that we could do the right thing at the right time.
Precise prophecy is impossible. Even so, the next best thing is to understand a company’s business options. Modelling consists of that looking ahead.
Establishing the likely effects of changes, be they ripples or earthquakes, enables planning well for bad times and better ones.
Our own example
For example, my company was hit by the coronavirus crisis just like almost everybody else. So, as soon as it became obvious that Britain would have a major emergency, we had to use the tools of our trade to prepare ourselves.
We started to make models for the size of the emergency, tracking expected national sickness cases, based on the figures available for the countries hit earlier. We constructed models for the economic fallout, and also divided our customers into three categories — first, second and third line companies.
- In this, first line meant companies in the tourist, restaurant and retail industry.
- Second line were those depending on big projects easily cancelled if panic and bankruptcies loomed.
- Third was everybody else.
We also altered our own ERP so that we could easily follow how much money was owing to us in each category, and how much of our working capacity was booked in each.
The day that we got our first call from a customer to say that they would shut down for seven months, we were ready. We knew what to expect, and prepared ourselves for a possible worst-case scenario.
Reasons for modelling
Modelling is no substitute for gazing into crystal balls! Modelling also needs frequent effort to maintain and adjust.
However, in times of rapid change, having a virtual representation of the future could mean the difference between survival or going down, success or failure.
We can build models to show us alternative future scenarios in important areas. Examples are: forecasting market size, managing stock, cash flow and working capital, production scheduling, profit-and-loss account, even valuing the company.
These can then show what to expect — almost creating the wisdom of hindsight but applicable to the future.
The secret of success in creating good models is threefold.
- First, you must have a good understanding of how things work in your company’s life.
- Second, you must have a competent software system to do the calculations. (This does not necessarily mean a fancy, expensive system — a spreadsheet is often sufficient.
- Third, a good balance must be found between software and human “grey matter”.
The model does crude quantitative work and humans qualitatively interpret and refine the results, aided by their experience, the latest information plus intuition.
Why modelling works
A business model helps managers to explore complex choices, developing a clearer understanding of inherent relationships between the variables and their likely outcomes.
For example, in manufacturing it’s customary to see the consequences of incoming orders for raw materials, parts, capacity and so forth. It is then a modest step to experiment with different likely scenarios for future eventualities.
- What will happen if the volume of orders increases or decreases, what effect will it have on cash flow, first in purchasing then in revenue?
- You can also foresee whether to arrange for extra work-shifts.
- If you need to purchase new machinery, or to stockpile materials — or whether just the opposite is the case, with a nasty slowdown.
If one sees the road ahead, preparing to turn the corner is easier.
In the services industry, capacity and cash flow are just as important. When a surge is coming, how long will it take to hire and train staff? When heading for a downturn crisis, is it necessary to lay people off or is it possible to finance a project or training which will use their work capacity to keep producing enough positive value?
A model also helps with all three stages of decision-making: analysis, choice and implementation.
First, you start by Analysing what you are up against
A business model will become a virtual representation of how certain parameters will develop. To enable this virtual world to be realistic, it’s necessary to include the most important influencing factors and make assumptions.
Factors like the business environment (for example, inflation, potential customers and taxes) and the organisation (for example, product prices, staff numbers and product volumes) need inputting appropriately.
Often, it is only during the process of building the model that it becomes possible to understand some of the complexities of our corporate life and how different attributes and external factors relate to each other. The purpose of the model is not to be 100% accurate, but rather to narrow the scene to a more manageable number of choices.
Then it is time to choose
There are two basic kinds of model, plus a combination of them. One kind of model factors in outside forces not controlled by us and sets out how we’ll have to perform. The other kind allows us to play with “levers” in our control and see the consequences of different actions.
A prime example of outside forces is the virus epidemic and economic slowdown of the country. We have no control over such events but it helps if we can get prepared.
An example for the “levers” kind of model is when we experiment with salaries, layoffs, financing internal projects and so on. These we have control over, and, after exploring different scenarios, the model will reveal potential outcomes so we can decide the preferred route.
Identifying the extent of this range of outcomes enables the model user to understand the potential risk and reward. The approval or rejection of the business choice is then made with the knowledge of the expected outcome and in the light of the risks that lie ahead.
Finally, Implement your choice
If the model resembles our real life operations then it can become the template for the execution phase, when we act on our choices.
Variances in the actual results achieved, against the ones predicted by the model, will help provide early warnings of unforeseen problems as well as enabling the project’s success to be measured.
The lessons learned from such monitoring should also help in planning the remainder of the project as well as improving the model.
Difficulties in building a model
While using models on a regular basis is often a game-changer, modelling is a sophisticated discipline and is not easy. Not everyone can do it — you may or may not have the mindset needed for modelling. Some people are born with this skill, others can master it if they try, but most people admit to finding model-building a source of confusion and despair!
Also, experience helps. Having built and used many models, you learn how to find the balance between an over-complicated model and an over-simplified useless one. To gain appropriate experience takes years, just like learning to play a musical instrument well, for example.
While it’s tempting to try and build the ultimate model, it usually does more harm than good. A highly complicated model will soon become cumbersome or impossible to update or modify.
Effective working models are usually rather straightforward and maintainable.
Some corporate software systems support modelling better than others. Generally, the dividing line is how closely the system’s operation and data structure follow life in the company.
It helps if the system contains some sort of simplified virtual version of the company — for example, in having a tiny virtual representation of every cost, worksheet, machine, worker or vehicle. Consequences can then be calculated on accumulated data for every resource item, showing where we are each day, calculating where we’re likely to be the day after, and then the following day, and so on. No human mind could do this with hundreds or maybe thousands of pieces of information, but computers can, over and over again for different scenarios, day after day.
Every model has its useful lifespan, and, as circumstances change, so do our needs for different models.
Models with a longer likely lifespan are worth building into an integrated corporate software system.
Other models, for short projects, can be better with spreadsheets.
The cost and time needed to create, adjust or maintain the model should guide which route to take.
Modelling is by no means only for crises and risk management. Indeed it helps us to find and manage opportunities as well. Once this crisis, like each one before it, will be over, we’ll have modelling to help us build a better (future-utilising) company.