Sales Operations: Forecasting

Ever since humans started the practice of agriculture, we have wanted one thing: consistency.

And yet this one, seemingly simple concept is incredibly difficult to obtain. Talk to any sales operations leader if you don’t believe me, because the primary goal of sales operations is not installing CRM, it is to maintain the balance of supply and demand: to sell exactly the amount that an organization is capable of producing. Talk to any development team leader who is not applying Agile, and you will understand the one thing they want is consistent, predictable development velocity. That’s why we have Agile; that is why every organization needs sales operations.

Accurate forecasting is therefore fundamental to the sales operations function and the entire organization. If you cannot forecast sales, then you cannot make any recommendations regarding supply, or finances, or recommend any actions to optimize profitability. If you can’t accurately predict the demand for your products and services, your organization is running through a forest, blindfolded.

So sales operations leaders are obsessed with increasing the accuracy of their forecasts.

And fortunately, there are a few game changers from the last decade, which can help you increase way more than accuracy. I am talking about timeliness and convenience, but I am also talking about solutions-focused forecasting, which empowers you to suggest actions that will result in higher revenues and profitability.

3 Reasons Why Sales Forecasting Is not All about CRM

I know that by “game changers,” you probably think I mean CRM. That is partially correct. Every CRM tool comes with a standard sales forecasting feature, which can be used to take sales forecasting quantum leaps ahead of spreadsheets.

But it is also true that the journey from spreadsheet to fully-loaded CRM functionality is arduous, and sales forecasting is an integral component of that difficulty, for three reasons:

1. Forecasting is completely unique for every market, and for every company comprising that market
2. Understanding your unique forecasting algorithm requires an iterative approach and will never be perfect
3. Accurate forecasting requires account executives to actually use the CRM in a predictable way

So, if a CRM vendor starts expounding on the accuracy of its forecasting algorithm, ask them if their engine understands that half your personnel use a funnel with 1 stage, and the others don’t have one.

If you are still using descriptive forecasting, then what I just wrote won’t mean anything to you. So I need to explain how CRM is powering predictive analytics, what that means, and how it makes consistency incredibly important.

Descriptive Versus Predictive Analytics

Plenty of companies use descriptive analytics to forecast sales because descriptive only requires historical data. The simplest example is to use the last two years of revenues. First, calculate last year’s growth rate, and then apply that growth rate to this year’s prediction.

In this example, the growth rate for 2017 is calculated by comparing sales in 2016 to 2017. But the forecast value is computed by assuming the same growth rate will occur in 2018.

This is a big presumption to make, and no matter how you slice it, descriptive forecasting just doesn’t tell the whole story. It’s like buying stock in a company because the company grew by 50% in the prior year. My question would be, why did the company grow by that much, and will that factor still be present in the year to come?

In other words, the problem with descriptive analytics is that it does not take into account the richness of factors that caused the historical data in the first place. And without that understanding, it really doesn’t do us a lot of good.

Predictive analytics begins to ask those questions, and it requires us to build a working knowledge of those factors which are the most influential predictors of future revenues.

What CRM Means for Predictive Forecasting

Any sales guy will tell you that in order to sell something, you need an opportunity.

So, if you want to predict revenues next month, then simply go and count up all your opportunities. You will be a huge step closer to predicting what your sales will look like next month. There is no doubt about it.

Predictive forecasting is incredibly useful in increasing the accuracy of your forecasts because it opens up the model to real-time data points.

But there is also an issue with predictive forecasting, which any real sales operations dork will know about: It is the optimization rabbit hole.

Predictive forecasting is a rabbit hole for this reason. Let’s say you have your opportunities model. Great, but now, one month, you buy a list with 100,000 emails on it, and you send out an email to all those cold leads. Are your sales going to increase exponentially? Your opportunities sure have.

Probably not, right?

So, you have to practice lead scoring, and weight each opportunity in accordance with its likelihood to close. The rabbit hole of predictive forecasting is figuring out your organization’s unique algorithm, what data points to include, and which will be more important than others.

Assigning weight, tweaking the model, and tracking results can require significant time from sales operations analysts, but it does deliver more accuracy in short-term forecasting. It also shows sales leaders how they can increase the likelihood of selling more, which can be a gamechanger.

It also takes a lot of time and effort, two things which CRMs are in the midst of solving right now.

What AI Means for Predictive Forecasting

CRMs like Salesforce and bpm’online have begun to apply machine learning to the forecasting conundrum. This solves the rabbit hole problem because all the hypothesizing and testing and validating is done by artificial intelligence, so your sales leaders don’t need to worry about algorithms. But they receive increasingly accurate predictions in real-time.

What this means is that the value of CRM is about to increase by a considerable margin. Einstein, Salesforce’s AI, is relatively new, so if you are still knee-deep in spreadsheets, you have some time before you will be engaging with Einstein.

But you need to start preparing for this, because AI-driven forecasting is only the beginning. We are heading toward a time when your CRM will be able to make data-driven recommendations to help sales leaders achieve goals. Because once an AI understands the leading indicators of revenues, it can tell you what you need to do to influence those indicators, and sell more.

3 Steps to Prepare for AI-Driven CRM

Here’s how you can prepare for that eventuality today.

1. Document your sales processes in customer-centric steps
2. Make sure that everyone knows and practices the same process
3. Increase consistency across sales teams