Everyone would love to be able to see into the future. This is especially true for management of fast-growing businesses.
Well, as far as I know, nobody has that power, but there is something that can get pretty close, and that is marketing attribution.
Generally, marketing attribution is known for looking back and not into the future. It takes revenue data and marketing data, and gives revenue credit to the marketing touches that led up to the sale.
While that is tremendously valuable in itself, marketing attribution perhaps contributes even more value by helping marketers use the integrated marketing and revenue data to accurately forecast future business outcomes.
Attribution solutions provide rich customer data that is the foundation for accurate forecasting. Specifically, there are two characteristics that attribution data provides that are necessary for accuracy: 1) data from the first touch to closed-won revenue, and 2) complete multi-touch representation of the customer journey.
Full-Funnel Data vs. Lead Data
Lead data only provides insights into one segment of the marketing funnel. However, without understanding the rate at which they converted to opportunities and customers, there isn’t enough visibility to make accurate predictions. Predicting the conversion from sales qualified lead all the way to closed-won based on lead data would be like predicting tomorrow’s weather in Seattle by looking at yesterday’s weather in Anchorage. There simply isn’t enough correlation.
But if you track data through the entire funnel, you’re able to make accurate predictions about how future prospects will flow through the entire funnel. When you can see exactly how every previous customer flowed through the funnel, it’s much easier to make predictions about how future customers will, too.
Multi-Touch Marketing Data
Additionally, multi-touch attribution improves forecasting accuracy because it is the only way for marketers to understand the complete customer story and properly classify them when making forecasts. Unlike in B2C, the B2B customer journey is long and requires many different touchpoints, which makes identifying the key marketing sources or channels not so simple.
When B2B marketers use a single-touch model, they put all their eggs in one basket, relying on one piece of information to represent the whole customer journey. Because of single-touch model’s inherent model bias, it makes it easy to misclassify leads.
For example, first-touch models overvalue top-of-the-funnel marketing activities and will tend to misclassify leads as such, when they may actually be better represented by their deeper-funnel touches or a combination of many different touches.
Multi-touch attribution allows B2B marketers to be confident that their customer data paints an accurate picture of how customers actually went from anonymous visitor to lead, and then from lead to customer because it captures multiple key engagements throughout the entire funnel.
With a multi-touch model, like the W-shaped model, credit for a customer gets split to properly account for the multiple touches that impacted the journey. By getting into the deeper granularity of the customer journey, leads, opportunities, and customers are accurately categorized to their marketing channel or source, which makes for more accurate forecasting. We will see an example of this later.
The Bizible 2016 Forecasting Process
When creating forecasts, the management team uses a right-to-left, bottom-to-top approach.
Right-to-Left
First, they ask themselves where the company wants to be in terms of revenue at the end of the period. We then figure out what it will take each month to get there. If we want X revenue in December, what will we need revenue to be in November, then October, then September, etc.? This gives us the revenue goal for every month.
Bottom-to-Top
Then, we work out the “bottom-to-top” portion of the forecast. This starts with calculating how many customers it will take to hit the monthly revenue number, and then continue to work backwards all the way up the funnel to leads.
Basically, it’s going through this series of questions: How many customers do we need to achieve our revenue goal? How many sales opportunities will it require to get that many customers? How many sales qualified leads will it take to get that many sales opportunities? How many marketing qualified leads will it take to get that many sales qualified leads? How many leads do we need to get that many marketing qualified leads?
If the marketing team only knows their impact up until the lead stage, it’s a big challenge to make these bottom-to-top estimates with any accuracy. They simply don’t have the right data to estimate the conversion rates for every stage’s transition from revenue to leads. But because we have full-funnel marketing data, we can confidently make those estimates based on historical data.
Not All Leads Are Created Equally
However, the “bottom-to-top” portion of the forecast has another wrinkle. Without attribution, every lead is the same. When working your way from the bottom of the forecast spreadsheet to the top, you would apply the same conversion rates to all leads, no matter the source or channel. However, we know that in reality not all leads are created equally. To assume that every lead will flow down the funnel with the same stage conversion rate would lead to an inaccurate projection.
Attribution irons this wrinkle out. With attribution, you can look at conversion rates by source or channel (or even something more granular) and apply more accurate stage conversion rates to your forecast, as applicable.
For example, we know that leads from events convert at much lower rates than from other sources. Let’s say your typical lead-to-customer conversion rate is 1%. But looking at your attribution data, you see that the conversion rate for leads from events is only 0.1% and events contribute 30% of your leads. If you don’t take the different conversion rates into account, your forecast will be in trouble.
Here’s an extremely simplified version of how that may play out in a 6-month forecast with a target of 90 customers:
Forecast without attribution:
What would actually happen with 1000 leads in January:
If you hit your forecasted lead goal every month, but 30% came from events, you would quickly notice that your forecast was inaccurate. By the time you get to June customers, you would be 24 customers short (27%).
Here is what your forecast would like if you had the same target customer goal (90 customers by June) but used attribution to do your forecasting.
Forecast with attribution:
When you take into account the lower conversion rate for event leads, you actually need 700 more leads starting in January to hit the same target customer goal in June.
These are, of course, made up numbers and many factors are controlled to isolate the impact of event vs. non-event leads, but it demonstrates how much your forecast can be off by when you don’t base it off of the right data.
When leads are properly classified and specific conversion rates can be applied, forecasts are a lot more accurate. This is enabled by multi-touch attribution.
With a smart attribution solution, the actual forecasting exercise of working right-to-left and bottom-to-top is exactly the same. However, because we use full-funnel marketing data and have the necessary granularity to look at how our specific marketing programs convert, the foundational data is sound and we are able to forecast with confidence and with accuracy. 