Thanks to Mack Grenfell and the team at Causal for this detailed post on building better revenue models. Read on π
It's not hard to make a case for your revenue model being the most important part of your overall financial planning and budgeting process. So many decisions are based off of your revenue projections (hiring, funding, capital expenditure et cetera) that getting your revenue model right should be a huge priority.
β
That said, revenue models can be hard to build. They depend on so many unknown factors - market & seasonal factors, product changes, marketing assumptions - that it can feel impossible to build a revenue model which you're happy to call accurate.
β
Whilst there are plenty of different factors to keep in mind when building revenue models, each of these factors is fairly straightforward when viewed in isolation. With that in mind, let's take a look through some of the steps and considerations you should take when it comes to building a solid revenue model.
β
If you're putting together a revenue forecast, then chances are that you're doing so to map out the growth trajectory of your business. Assuming that your business is growing, the main things you'll want to consider in your model are:
β
β
β
Save for the few lucky businesses that can grow purely organically, most businesses need marketing as a way to fuel growth.
β
A naΓ―ve way to include marketing projections into your revenue model is to understand the cost per acquisition that you get on each of your marketing channels, collate estimates for how much you're going to spend on marketing in each future time period, and use this to forecast user growth.
β
I call this method naive as it only really works if your future marketing spend is very similar to your current marketing spend. If this isn't the case, then your current cost per acquisition figures are likely to be under-estimates of how much you'll have to pay to acquire customers when you're running with higher marketing budgets.
β
You might consider using regression analysis or econometric techniques like media mix modelling to understand how your cost per acquisition figures are likely to scale with marketing spend, if you're planning on changing your spend from current levels.
β
β
β
Whether your business monetises users through distinct purchases, monthly subscriptions, or through usage, it's critical to have some sort of user LTV (lifetime value) model in place to feed into your overall revenue model.
β
Having an LTV model (even if fairly basic) helps you to understand how much value you can expect to extract from your current users throughout the rest of their life-cycle.
β
If you have the resource to really dive deep on LTV, you should consider approaching this in a cohorted manner. That is; breaking your current user base down into monthly or quarterly cohorts, and using your LTV model to estimate how much more revenue you'll be able to generate from each cohort over its remaining lifetime.
β
A cohorted view of users is great as it lets you:
β
β
β
β
β
There are virtually no businesses whose revenues aren't subject to external factors like seasonality.
β
This is simple to understand for businesses selling low-value consumer products, where changes in demand throughout the year affect your revenue. It's important to realise though that seasonality plays a part even if your revenue comes through large enterprise deals; the likelihood of these deals completing will vary by time of year.
β
The best way to account for seasonality is to look back at previous years' data, and to apply patterns from those years onto your future projections. For example, you might calculate your baseline level of revenue in a historical year to be $X, and run some analysis to determine seasonality based on how far above or below $X of revenue came in each month.
β
Of course if you've grown through all your businesses' previous years, you'll likely be better off trying to understand seasonality in terms of how it's influenced revenue growth rates, rather than just pure revenue.
β
This data can then be layered onto your revenue projections going forward, to give a more accurate intra-year forecast. Of course, factoring in seasonality shouldn't change your revenue projections on a yearly view; rather it should precisify your quarterly/monthly/weekly revenue projections.
β
Even if factoring in seasonality might not seem like it affects the big picture, it's still worth doing as it can give you a far more accurate view of cashflow, which in turn allows you to plan your expenses with greater confidence.
β
β
β
Having a solid revenue projection is great, but it's important to realise that it usually just attempts to forecast a median or average-case scenario. With so many assumptions going in to even a basic revenue forecast, you should never expect your revenue forecast to be totally accurate.
β
This then begs the question - just how accurate is my revenue forecast?
β
If you know with 90% certainty that your revenue will fall in some narrow band around your projections, this can give you a lot of confidence to plan expenditure around your forecast. Conversely, if there's a non-negligible chance that your numbers could come in way off your forecast, then you ought to respond by being more conservative with your budgeting.
β
β
In order to understand the degree to which your actual revenue figures are likely to deviate from your projections, you need to build uncertainty into your model's input assumptions.
β
Rather than assuming that your new user growth is going to be exactly 5% of your current user base each month, look back at historical data and see how accurate this assumption is. If, in 90% of historical months, user growth has been between 3-7% of existing user numbers, then replace your 5% growth assumption with a probability distribution that lines up with your historical data.
β
Once you've gone through and quantified the uncertainty in your assumptions, you can then use modelling techniques like Monte Carlo simulation to stress-test your models, and understand just how confident you ought to be in your top-line projections. This can then feed through to the rest of your financial planning, and give you confidence that your budgeting will hold up even if your revenue projections aren't precisely on-target.
β
π€Ή Track and forecast your marketing costs
β
πΏ Create a LTV to estimate your overall revenues
β
β Keep track of users' seasonalities
β
The above are just a few ideas for ways to make your revenue projections more robust. There are always more details to add, and many will depend on the intricacies of the specific business whose revenue you're trying to forecast.