What is bottom up revenue modelling

Overview

Bottom up modelling provides a window into future business performance that cannot be attained using top down modelling, and requires a heavier lift to build out the appropriate revenue models and spreadsheets for planning.

Working upwards from marketing and sales efforts (activity & leads), bottom up modelling follows the flow of activity, leads and revenue through the business to give an understanding of the end output of current GTM efforts.

The calculation flow is simple, using the volumes at each stage and their conversion rates to understand how much passes along the funnel - Where both outbound sales and inbound marketing contribute towards sales opportunities, you'll find that multple funnel outputs converge to input into another.

With bottom up modelling however, you may find that the possible output of your bottom up model does not reach company revenue targets, and this is where a top down bottom up approach comes into play.

Bottom up revenue modelling uses a multitude of different revenue formulas (and so for a working example it's easier to watch the video), however the overall approach is as follows:

Top Down Bottom Up

More of a bottom up, top down approach, this approach uses a bottom up model to find the gap and a top down approach to calculate what is needed to plug the gap. Like with other top down models however, the top down aspect can work off incorrect assumptions, meaning that it then needs to be rechecked using a bottom up approach again.

This iterative process works through making changes to the plan until project revenue meet the company targets.

This approach to modelling is also far more applicable to scenario testing, which helps you to understand the implications to changes to any of the metrics/data points that exist in your model, helping you to build out additional iterations of the plan to meet each scenario.

Whilst this approach to planning provides a far more realistic and useable approach to projecting the longer term revenue output of a team and its growth plans, it's is not only intensive (taking businesses anywhere from weeks to months to build out), but can become stale fast needing the most recently updated data to remain as an accurate approach to revenue projection.

Building a bottom up model

In our example, building up from the bottom up begins with 1,000 marketing leads (defined as MQLs in this example), and a revenue target of $10,000,000 ACV.

It's worth pointing out that the conversion rates in the example are for easy maths, and should not be taken as a benchmark for your own performance!

Through understanding the conversion rates between key stages of MQL > SQL > Close, we can understand how the flow of leads converts into sales opportunities and then onwards to won deals.

Through knowing that MQLs convert into SQLs at a rate of 45%, we know that 1,000 MQLs convert to 450 SQLs.

The same approach is applied tor SQL to Closed Won deals of 33% (actually 1/3 but we hid the rounding). That means that the 450 SQLs convert to 150 sales.

In our example, we have the average ACV as being $50,000. With 150 deals in the bag this gives a revenue total of $7,500,000, leaving a large deficit. With the deficit in hand, it's time to work backwards down (back to the top down!) to understand what routes are possible to close the gap.

By dividing the deficit for $2,500,000 we know that we need an additional 50 deals to close the gap, and by dividing those 50 deals by the conversion rate of 1/3 we can calculate how many sales opportunities we need to generate as SQLs.

From here we can then work back and understand the amount of marketing leads we would need to generate, or the amount of outbound SDRs.

The video shows you working back to both of these routes, as well as comparing different cost & attainment scenarios that could lead an organisation in each of the two routes.

The important thing with bottom up modelling is to understand that it's a more intensive and iterative process that needs joining up with considerably more data, as well as an understanding of overall & utilised capacity.

Bear in mind that the above worked example is for a single snapshot point in time - You should ideally be modelling this over a 24-36 month period to be able to build a longer term picture of the decisions and strategy calls that you need to make over the next 12-18 months.

Supercharge this approach with live data and active scenario testing and you can have an approach that's a constant understanding of where your business will land at any point in time, you might struggle to achieve this in a spreadsheet however...