One of our customers mentioned it in passing. "Funnelcast is like Moneyball for B2B Sales." We kind of like that moniker and thought we should explain what they meant.
If you haven’t read the book or seen the movie; Moneyball chronicles Billy Beane, general manager of the underdog Oakland Athletics.
Beane (Brad Pitt) adopts a radical and (at the time) controversial approach to making Oakland a top tier competitor in Major League Baseball (MLB). Challenged by his limited budget, Beane builds a team of undervalued players by taking a data-driven approach to identifying, analyzing and signing talent.
"Any GM that doesn't tear down their team and rebuild it using your model is gonna be a dinosaur."
~ Red Sox owner John Henry to Billy Beane, as portrayed in Moneyball
In the 2002 season, Beane and the Oakland A's went on to win 20 games in a row, upending a long-standing paradigm in MLB where well-compensated scouts used their gut-feel to inform the decision making of general managers.
Wondering what baseball has to do with sales forecasting and sales optimization? Plenty!
Like what Moneyball did for the Oakland A’s, Funnelcast makes it possible to compete more efficiently through data-driven sales forecasts and sales optimization.
When we say sales forecasting, we are referring to the ability to make predictions based on historic performance, the current opportunities in your funnel, and future opportunities not yet in your funnel. By sales optimization we mean using forecasts to make resource allocation decisions or to improve your sales processes. We sometimes therefore use the terms sales optimization and resource allocation interchangeably. Long-term forecasts are the building blocks of sales optimization decisions.
So, what did our customer mean by saying that Funnelcast is like Moneyball for B2B sales? Let’s look at two examples.
A forecast can compare the productivity of opportunities in multiple business segments. If opportunities in business segment A are worth 5 times segment B; you might use that information to reallocate resources to segment A from B. Similarly, your win rates by stages shows you that you have a large gap between two sequential stages, that’s an indication that you should probably optimize your sales process with a new intermediate stage that tells your sales team what to do next to advance opportunities.
DATA BEATS INTUITION Humans are prone to cognitive biases, or errors in judgement due to personal beliefs. In Moneyball, the scouts hired to identify up and coming talent exhibited a number of biases, including overconfidence, herd mentality, confirmation and hindsight bias. Beane starts challenging his scouts with data to find talent; resulting in a spectacular season for Oakland and the beginning of a paradigm shift in MLB.
Sales resource allocation based on intuition often misses the mark, due to cognitive biases, just like the baseball scouts. “I know our sales pipeline solid”, “we’ve always done it this way”, and “that's a 50-percenter” are all manifestations of biases that impact forecasting and resource allocation.
We are not talking about what a salesperson is able to forecast in the last weeks of a sales process. It’s possible to get a really good forecast of an opportunity just before it closes because the salesteam is actively engaged in discussions with the prospect. There is a lot of active feedback at this point.
But most salespeople don’t make long-term forecasts or receive feedback on them. They can’t develop an intuitive feel for how opportunities will resolve over the long-term. Salespeople also tend to be optimistic; and any feedback they get is so decoupled from the earlier forecast that the feedback/improvement effect is miniscule. Moreover, long-term predictability is just generally poor because the signal is weak.
This is precisely the area where computers beat people. People are not very good at picking up on weak signals—particularly when the response lags the signal by a lot, as in sales forecasting. But computers can analyze the entire history of sales experience and will outperform individual judgement at finding trends and differences to exploit.
Data-driven forecasting not only removes human biases, it often leads to counter-intuitive insights that might easily be rejected when decisions are made by gut feel. It’s exactly those unexpected outcomes that often lead to a complete rethink of sales team staffing and resource allocations, and can make a significant bottom line contribution.
SMALL DATA/BIG DATA—YOU'VE GOT THIS In the era of big data, readers might think they need large volumes of sales data to benefit from data-driven forecasts. But what constitutes "big data" depends on your perspective. (Learn more about the sales forecasting Big Data myth here.) Big data for a human is tiny for a computer. Much like MLB, where a full roster is only 26 players, all but the largest corporations have modestly sized sales teams, generating only a handful of metrics for each team member and each opportunity.
Funnelcast makes efficient use of data that is automatically recorded by CRM systems to generate accurate forecasts. This data may be the history of hundreds to thousands of opportunities over several years. It is way too much for a human to synthesize intuitively. But for computers, this is small data—easily processed and analyzed.
While more data will improve forecasts, Funnelcast typically makes realistic forecasts with as little as three sales cycles of data. So if your sales cycle is six months you need 18 months of sales history. This history is automatically recorded in most CRM systems.
A key component of data-driven forecasting is feature engineering, which transforms raw CRM data into a format that enables machine learning to effectively separate actionable signals from random noise. In baseball, data analysts discovered features that proved far more predictive of player performance than the traditional batting average metric. For example, on-base percentage is the total number of hits + bases on balls + hit by pitch divided by at bats + bases on balls + hit by pitch + sacrifice flies. While simple to calculate, this feature is not as easy to comprehend as batting average, and might easily be rejected in gut-based decision making.
Most important, this feature engineering process is actually an unbounded problem, and this is where many analytic modelers get stuck, endlessly testing new “features” only to determine if they improve forecasts (they usually don’t).
Funnelcast comes preconfigured to use features derived from standard CRM systems, which is why you can expect to see meaningful and actionable insights within minutes of connecting your data to Funnelcast.
CONTROVERSIAL BECOMES MAINSTREAM While Moneyball is about baseball, it’s also a fascinating story about the inherent difficulties of changing an organization’s culture and business processes. The ability of scouts to identify talent was rarely questioned, until Beane began using data to challenge that assumption. It turned out that many scouts barely outperformed random selection, certainly not justifying their lucrative paychecks. The truism that “data doesn’t lie” often takes a while to produce organizational change, whether you’re running a baseball team or a sales organization.
TIME FOR A NEW PARADIGM IN SALES FORECASTING While practically every sales organization collects metrics and data in a CRM, utilizing that data effectively to drive accurate forecasts and realistic resource allocations still eludes many firms. Funnelcast offers an automated, turnkey approach to making sense of sales data, leveraging the latest machine learning technologies. Much like Billy Beane’s heterodox approach to player selection upended decades-old practices built on gut feel, sales forecasting is ripe for change. Let Funnelcast show you how your sales data can generate better insights to inform your sales process.
Want to learn what else your CRM data can do for you? Check out part two of this blog.
Photo by Jose Francisco Morales on Unsplash