Top-down vs Bottom-up Forecasting: What You Need to Know

Predicting your organization’s financial health through forecasting is a critical role of your finance function, but there are many approaches. Two of the most common methods pitted against each other are top-down vs bottom-up forecasting.

Both of these have their strengths and weaknesses. Your choice of approach might depend on your company size, market share, how much time you’re willing to spend on forecasts, and your strategic objectives. And in fact, you might not need to choose at all.

In today’s post, we’ll break down the pros and cons of top-down vs bottom-up forecasting and when to use each.

What is top-down revenue forecasting?

In a nutshell, with top-down revenue forecasting, a company assesses the size of the market it operates in and its share of that market. For example, with a top-down forecast, a company like Salesforce may calculate its forecast revenue based on the total value of the CRM software market and its share of that market.

For organizations with a large/dominant position in their market, this isn’t an unreasonable approach to take.

A slightly more nuanced approach to top-down revenue forecasting when you already have a historical track record is to just project that forward: if you did $1, $1.1, $1.2, and $1.3 million in each of the last 4 quarters, and there’s no obvious change in the market dynamics, it might be reasonable to assume about a $100,000 gain per quarter going forward over the next few quarters.

What is bottom-up forecasting?

On the other hand, bottom-up forecasting involves analyzing the individual operating drivers of the business and building a model that translates those into a revenue outcome.

These drivers include inputs like:

• Marketing spend to generate leads.
• Conversion rates from leads to marketing qualified leads, sales qualified leads, opportunities, and won deals
• Net dollar retention that factors in upsell and churn on existing customers

In simpler terms, top-down models start considering the entire market and derive forecast revenue based on market share. They don’t make any particular assumptions about how your business works. In contrast, bottom-up forecasts begin with the individual business drivers, and build up to a forecast revenue figure, but don’t really factor in market conditions. 

Understanding the pros and cons of both these types of revenue forecasting and when to use each is key to supporting business growth.

 

 

 

 

 

The pros of top-down forecasting

The biggest advantage of taking a top-down approach to revenue forecasting is that it’s quick and easy. That’s because organizations don’t need to do a deep dive into every single aspect of their activities to create a forecast. This can be a huge time- and resource-saver.

Additionally, a top-down revenue forecasting approach requires little or no historical data. As EY points out:  “past sales data is often not available for startups as they might not have launched yet and are thus not generating sales. Using the top-down approach can thus be a quick win as it is based on market data, which is often more easily available for pre-revenue startups than sales data.”

Top-down revenue forecasting also provides a high level sense of what is possible in the market. We’ll come back to this point later. And because the approach is quick and easy, if you’re operating in a dynamic environment, or you’re an early startup with a lot of potential options on the table, you can reforecast often, or forecast against multiple possible options, based on new market data to update your thinking on the potential for your business.

The cons of top-down revenue forecasting

Top-down revenue forecasting has serious drawbacks – especially for companies in growth mode. Its biggest one is that it can be overly optimistic/inaccurate, as it is based on broad brush assumptions rather than a plan to achieve a target. 

While an optimistic forecast may draw investor interest, they’ll want to see a credible operational plan for achieving it.

Top-down forecasts are also difficult for smaller companies to use — especially if they’re operating in a large Total Addressable Market (TAM). For a leading player in a large industry, it’s relatively easier. For example, if Microsoft Azure has 22% of the cloud market in 2021 Q3, it’s reasonable to assume it will have a similar market share in Q4, and take the expected growth of the overall cloud market and get the dollar figure for Microsoft’s share.

If SmallCo has has 0.01% share of the cloud market, a small fluctuation in market dynamics could lead to a very different outcome in Q4. A top-down revenue forecast also doesn’t reflect the levers the organization has control of, like the number of sales reps they hire or their marketing spend.

Finally, top-down revenue forecasts aren’t operationally very useful. When growth exceeds or misses the forecast, the first question is why? If you don’t have a forecast model that connects how your business runs to the revenue it generates, it’s really hard to answer that question.

The pros of bottom-up revenue forecasting

Forecasting revenue bottom-up gives you a more accurate, detailed, and useable tool. It connects the organization’s data and structure and explains how its activities can impact its financial health. As a result, bottom-up forecasting helps you make data-informed decisions.

Bottom-up forecasting is also useful because it’s driver-based. That means it focuses on helping organizations understand the drivers, or key factors, that go into hitting key results and the relationships between them. This allows businesses to better understand what’s behind their success or failure and pinpoint what needs to change if things don’t go to plan.

A bottom-up approach is also a more collaborative process than top-down forecasting. It provides more opportunities for departments and key stakeholders to share their unique perspectives on underlying issues or ways to improve a company’s bottom line.

The cons of bottom-up forecasting

Given how detailed it is, bottom-up forecasting can be much more time-consuming than top-down forecasting. And sometimes, without the right tools, by the time a company collects and processes information for a bottom-up forecast, it could be too late to use it. 

Additionally, any errors a company makes (like an overly optimistic close rate) at the micro level are amplified as they approach the macro level. Errors often compound as time goes on — this means FP&A teams should regularly challenge and update assumptions for more accurate forecasts. This is especially true when a lack of data forces modelers to make more assumptions. For example, a company that’s just found product-market fit doesn’t really know what price it can charge, what customer segments will be most amenable to its offering, what it’s CAC will be. It likely doesn’t have good CRM data to be able to measure its sales cycle or the number of customers it will have one, three, or 6 months into the future. A great modeler might have enough experience to make reasonable estimates based on industry norms, but it’s almost certain that the current business won’t fit those norms exactly, especially early on.

Finally, with bottom-up forecasting, departments can sometimes deliberately provide inaccurate projections that sandbag the model and their targets so their results exceed expectations.

Top-down vs bottom-up forecasting? A hybrid model may be the best option

Forecasting sales is good; proactively making sure that the business can execute against a sales target is better. The best revenue forecasts incorporate elements of both top-down and bottom-up approaches because they can serve as a check on one another. A top-down forecast can serve as a check to make sure that a bottom-up forecast is realistic. And a bottom-up forecast can illuminate the operational requirements that need to be in place to hit the top-line goal.

Whether it’s top-down, or bottom-up, the best way to approach forecasting is to use a rolling forecasting model, which you can learn more about in our guide:

A step-by-step guide to rolling forecasts

***

Ed note: Want to dive deeper into essential FP&A topics? Check out OnPlan’s most popular posts of all time here:

• SaaS profit margins: how to calculate them and what are the benchmarks?
• An Introductory guide to rolling forecasts 
• Sales forecasting in Excel: How to do It + free template download
• How to use a headcount planning model to determine your next office location
• Best practices for building a remote-first workforce planning model
• Bookings vs billings vs revenue: What’s the difference?
• What does the SaaS magic number mean for business growth?
• How CFOs can prepare for the Great Resignation