It can feel intimidating when the actual numbers start rolling in. Your predictions are being put to the test and you want to be as right as possible.
But it’s not about getting everything right the first time.
When real-world data starts coming in, don’t think of it as proving the model wrong or right. Think of it as an opportunity to learn, improve, and present results.
Timeliness is the name of the game with forecasting and testing projections. You want to catch any changes to the trends before they disrupt decision-making, present outdated information, or impact financial performance.
If you can be prompt and present the right information, you’ve unlocked the code to provide high-value projections and planning.
Even if you predicted everything correctly down to the last cent, you’d probably think something went wrong. Unless you’re working with extremely simple financials, some deviation is part of the game.
Variance isn’t necessarily a bad thing. Explainable variance further informs your model so long as you understand why any predictions or assumptions were off. If it’s unexplainable, then you’ve got a problem.
As the real-world numbers start rolling in, it’s your job to explain the differences, learn from the outcome, and iterate on your model. Explaining the differences helps your client learn about their business, and by improving your model, you’ll get more consistent, reliable results down the line.
Throughout this series, we’ve talked a lot about drivers. Essentially, drivers are the factors that most impact a business’s outcomes and decision-making.
For example, a business that manufactures goods would want to understand how their cost of goods and pricing impacts their gross profit and optimal sales number. On the other hand, a service provider cares more about customer retention.
You’ll have assumptions set about what these drivers are. Now that the real-world numbers are coming in, it’s time to test whether that’s true.
Try to find the correlation between different revenues, expenses, and operational data. Every month, you want to check whether what you assumed drives the business had the expected effect. If not, what was that driving factor?
Don’t forget to look at external factors that could throw your projections off. Just because your prediction was off doesn’t mean you should throw the baby out with the bath water. Chat with your client and try to dig for information that’s outside of the financial reporting that may have impacted the results.
Correlation isn’t causation, but it’s still important.
Part of understanding how a business works is understanding how a change in one part impacts every other aspect of its operations. With financial data, you get insight into how a decision in one facet affects the outcome in others.
Start with straightforward starting points like finding the correlation between marketing spend and revenue.
As you start to connect different metrics together, timeliness becomes integral to your services. With these connections, you can start to infer where the business is heading based on a trend in the numbers.
You can become a pre-emptive signal of what’s coming up for the business. Turn early trends into action by outlining how those stats have been connected to success in the past. The more that you can connect trends in a client’s financial reporting to real-world actions, the more you create value for your client.
A perfect example of this is predicting the necessary head count. If you see that the data is showing sales and production trending up, you can report that your client needs to think about hiring. Given that hiring is a long and involved process, this helps them get the staffing level they need before it becomes a problem.
The other part of working with the actuals is measuring the outcome of certain decisions.
With driver-based modeling, there are specific metrics that your client will use to guide their decision-making. Now, it’s time to break down whether the decision had the intended effect.
When working with your client to analyze the numbers and start shaping decisions, always try to have a measurable, expected result you can track progress against. This is how you’ll decide whether your expectations were correct.
If the outcome wasn’t as expected, look for reasons why. Can this new information reshape your model? Your client will appreciate if you learn from the experience and use it to improve your projections and models. Frame it as “The result wasn’t expected because of this factor and I’ve updated the model to reflect that.”
Iterate as often as you can. The longer something goes unnoticed, the bigger its effects become. If there’s a disruption in the model, the sooner you catch it, the lesser the financial impact. Financial modeling software applications can help you set up forecasts and make changes to your models quickly and easily, which enables you to provide impactful and timely advice to your clients.