Profile picture for user Alexis Getscher

How many times have you managed marketing in Excel? 

What B2B marketers try to answer with average conversion rates and forecast models in Excel, often leads to plans that are time consuming to create, unwieldy, and lack forecasting accuracy. 

With the advancements in martech today, marketers are able to move out of Excel to accurately forecast and plan for revenue. Here’s how to do it.


What is Marketing Forecasting

Marketing forecasting is the ability for marketers to showcase the downstream impact of their efforts. For example, if I publish an e-book in January, I’ll get x amount of leads in February and x amount of opportunities in March.

Advanced marketers will forecast down to revenue. As in, if we attend an event in July, we can expect x leads in August, x opportunities in September and x new customers for x amount of revenue in October.

To do this well, marketers must know historical conversion rates for each channel and each type of content to determine how leads will flow through the funnel. (i.e. e-books convert leads at a different rate than blog posts). Marketing forecasting tends to be unreliable because it’s based on averages across multiple channels and metrics.

The Flaw of Averages in Marketing Forecasting

To forecast, marketers often rely on averages. The problem with this approach is that too many assumptions are made, and those assumptions get exponentially more inaccurate the farther out you forecast.

average forecasting

For example, marketers take the expected number of leads in each channel and multiply that by the average conversion rate per channel. Then they sum all the average numbers from each channel to get the total expected opportunities for that month. From there, the expected number of opportunities gets multiplied by the average win rate and average deal size to finally arrive at forecasted revenue.

The problem with this is, every lead is different and there’s a lot of  variation across leads in regards to when they will convert and how much the account’s deal size will be. This approach ends up being the average of the average and it’s highly inaccurate.

Instead, marketers should take a granular approach.

A Foundation of Good Data

For simple forecasting methods to produce good forecasts, they must use data that is low variance and has predictable trends. And we all know that marketing performance is often unpredictable and follows complex seasonality or trends. Data that varies highly, means that simple forecasting methods will likely produce estimates that are not close to actual results. If each channel has these high errors, forecasts will be inaccurate and marketers will miss the mark on their goals.

forecasting errors

Accurate marketing forecasting depends on good data. Without a starting point of accurate data, marketers are basing their projections on assumptions. For data that has complex, underlying structures such as seasonality and multiple causal factors, marketers need to use more advanced forecasting methods to minimize the magnitude and variance of the forecast errors.

 For that reason, to accurately forecast, marketers must have a multi-touch attribution solution in place.

Attribution allows marketers to start from a granular position, instead of a position of averages. With Bizible, all online and offline touchpoints are tracked for every lead and that data is then pushed into a predictive model based on machine learning.

predictive forecasting

This model then forecasts revenue for every opportunity, based on specific data for that opportunity. No more averages.

Marketers can then define, segment and forecast across stages. Which leads to our next section: marketing planning.

Present a Menu of Accurate Alternative Marketing Plans and Strategies

When forecasting begins from a position of granular data and connects to revenue, it makes it possible for marketers to plan for the future. 

With Revenue Planner, CMOs can now see the future revenue impact of current marketing investments. For example, invest $10k in social in July, what’s the revenue impact in August? What about September or October?

The machine learning model takes into account seasonality, sales cycle, and touchpoints data, and makes it possible to plan across multiple channels simultaneously.

It gives CMOs the power create marketing plans in a single, accurate platform, instead of trying to combine and make sense of aggregate data pulled into Excel from multiple platforms.

Just like how financial advisors present multiple growth and investment strategies, CMOs can now present the CEO with various marketing investment plans. For example, Plan X with lowest risk and steady growth or Plan Y with fastest growth, but highest risk. All plans can be reviewed and discussed with leadership team to decide the best option for the business moving forward.

Marketing forecasting and planning no longer needs to rely on averages and assumptions, marketers can now accurately invest in the future. It’s time for CMOs to join the rest of the C-Suite in planning to revenue.