Forecasting for Uncertainty

As a CFO and as a business owner, planning is at the heart of everything I do; ensuring the business has sufficient cash to meet its obligations, understanding when it makes sense to bring in new people, not so early that we’re just carrying cost, not so late as to have negated the need and lost the opportunity, what are the best revenue streams for the business and where are we going to get the best returns. The greater the degree of certainty the easier it is to plan.

Out in Papua New Guinea where I lived for close to a decade, life was uncertain and for many survival was the only goal worth pursuing. The business environment was just as unpredictable, a landscape of fat and famine and economic cycles that turned on a coin.

I’ve produced and worked on dozens of financial models for Boards, shareholders, investors, financiers and management. They all had one thing in common. They were all wrong.

Because a forecast or financial model is not a budget, it’s an illustration of possible future performance. It’s an exercise in structured thinking that allows a business to run a thought experiment on ‘what ifs’ and to make some educated assumptions on the outcomes. The more stability and the more history a business has, the greater the level of reality the model will reflect.

Businesses that traditionally operate under a high level of uncertainty and risk (high growth businesses or startups, businesses operating in non traditional markets or with new products) are those that will have least ‘accuracy’ in their numbers, but counter intuitively this is the reason they must have gone through the greatest amount of rigour in the modelling process.

Here are some of the most important steps to building a useful financial forecast or model.

Excel at Excel

It is pointless trying to build a decent financial model if your Excel / Sheets / whatever skills are weak. You won’t be able to build either the complexity, or the flexibility you need to into your model and, believe me, you’ll want to. Find someone in your organisation who can — if it’s the person who manages the business’ finances then so much the better.

Understand Your Audience

I’ve mentioned this before in other contexts. It’s imperative that you have a clear picture of who your model is intended for and why. Banks focus on different information than investors, heads of department need different information than the Board.

Talk to Your Team. Then Talk Some More

Before a single number gets entered you want to be gathering information from every corner of your business. If you’re a one man band then clearly that’s going to be a little easier than if you’re heading up a multi-divisional organisation. This is effectively a variation on Steve Blank’s exhortation to ‘Get out of the building’; information is not static and is definitely not centralised. For the most part it resides in people’s heads, you’ve got to find a way of extracting it.

Question Everything

If you don’t, believe me, someone is going to. Do we need to purchase that asset in three months, can we wait till 3rd quarter? Is that hire too junior, if we’re going to release product on spec and on time do we need someone who can hit the ground. Don’t underestimate, even in a small organisation, individuals propensity to protect — their budget, their department, their arses.

When You’ve Exhausted Internal Knowledge, Really do Get Out of the Building

Or jump on the internet, talk to suppliers, to customers, read a paper or two; do quality research. What is the macro environment looking like, is there legislation that’s going to influence your business, does Brexit mean that your non sterling denominated payments are going to be higher and cash more scarce, does Trump’s wall mean fewer exports for you or more sales to service the guys building it. Like is said right at the beginning, the more certainty, the more your model’s accuracy improves.

When it Comes to Tabs — More is Definitely Less

I’ve seen models with one tab and all the numbers entered on to it, no formulas, no detail. Bad. Sad. Things change. They become more certain, they become less certain. You want a model that you can keep coming back to, keep adjusting. In the last multi product, multi regional, multi entity model I built, because we had a fairly static product suite, I had revenue and cost of goods on separate tabs for each year. My HR tab was a staff sheet that took on incremental costs and allowed for inflation. It fed into a number of additional sheets that allowed me to quickly and easily adjust for part time or redundancy. My overheads for product and travel sat on separate tabs because we had detail, everything else was on a formula driven list that made it easy to add spend in where necessary. The more you break up your business in this way, the more quality you will bring to the model.

Alternative Facts

As I said early on, financial modelling is a way to illustrate the shape of a business and its potential futures. This is the only time where I would advocate the use of alternative facts. These are the levers that drive your ‘what ifs’ — revenue down by 30%, I need 2 in-house developers and I can get rid of my outsourced tech development, if I had product x in territory y this is what the business would look like. All of this helps to drive better decisions and for you to work with your key stakeholders (especially at the Board and shareholder level) to drive the direction of the business and to better understand the holy grail of any business operating under uncertain circumstances — how much funding you need till you’re cash breakeven.

Make no Assumption

That is, make no assumption that anyone not involved in drafting your model is going to understand it by just picking it up and going through it. List all your assumptions so that it’s easy for others to understand your thinking, highlight cells that aren’t formula driven and where input can be changed so others don’t break your spreadsheet but can still play around with it.

What Else?

There are plenty of other things you’ll need to work out, how to summarise the key financial data, how detailed your balance sheet is going to be, what format to show your cash flow in, how to show departmental splits, consolidate other entities, what exchange rates to use, what inflation to build in. The one thing that I think is critical is to incorporate the most important non financial KPIs that users of the information are going to want to see.

Make no mistake, forecasting in uncertain environments is a complicated undertaking, it needs skill, a logical mindset and an eye for detail — not to mention a healthy dose of imagination!