Customer Success

Advanced Sales Forecasting Methods: Getting More from Your Sales Forecasts and Improving the Quality of Your Pipeline

August 13, 2012

Sales Forecasts
Editor’s note: This guest article from Swayne Hill, co-founder and SVP Global Field Operations at Cloud9 Analytics, builds on basic information provided in OpenView’s eBook Sales Forecasts: A Question of Method, Not Magic and provides additional steps for taking your sales forecasting methods and efforts even further.

Sales forecasting is primarily a management reporting exercise for the CEO/Board, but if you’re not leveraging this process to make real, impactful improvements to your business you’re missing a huge opportunity.

What if I told you your CRM system could identify hidden risks lurking in the pipeline before you submitted the forecast?  What if it could alert you to pipeline coverage issues looming on the horizon? It can – as long as you know what an ideal sales situation looks like (proper prospect profiles, an appropriate level of rep engagement, etc.) and you’re willing to make a couple of tweaks to the system.

Here are three steps you can take to get more out of your sales forecasting and improve your overall sales pipeline health.

1) Separate ‘Forecast’ From ‘Win’ Probability

What in the world does an Opportunity Probability mean anyway? Is it the probability this Opportunity will convert to a win?  Is it the probability this Opportunity will close on a certain date? What about the probability the Opportunity will be won for the specified amount? Perhaps it’s all of the above. But that’s exactly the issue — it’s too confusing. When you’re developing a sales forecast or managing risk out of the underlying sales pipeline, confusion is your enemy, so let’s get rid of it.

Have your CRM Admin change the name of the standard Probability attribute to ‘Forecast Probability’. This will represent the Sales Rep’s opinion of how likely it is a deal will close by a certain date for a specified amount of money. As the sales cycle advances, so too will the Forecast Probability. It’s usually linked to Sales Stage but can be overridden by the Opportunity owner. Remember, though, it’s subjective, which means while it may be useful for holding a sales rep’s feet to the fire, it’s not so good for calculating an accurate sales forecast.

Next, add a new custom attribute to the Opportunity record called ‘Win Probability’. This is a locked calculation which will evaluate any time the Opportunity object is updated. The formula will be based on factors that have a high-degree of influence on win/loss outcomes. Leave it undefined for now — we’ll determine the detailed calculation below. Make sure not to confuse Win Probability with Win Rate (which you’re likely already tracking). Win Rate will be the number of deals you win divided by the total number of closed deals for a specified time period. Win Probability, on the other hand, is a somewhat predictive indicator of the likely outcome of any particular deal. The two are related but there are no dependencies – you need both for different reasons.

2) Create a Sales Opportunity Scoring Model

Before moving forward you’ll need to determine the influential characteristics that have the highest correlation with winning or losing a deal. For example, do you normally win with small companies in the publishing industry where the sales person engages a ‘student of the game’ finance director and all requirements are validated by the line of business end-customer of your solution? What if the deal stays too long in Stage 3, or there has been no contact with Finance at all? Do these represent risks? I bet they do.

In the example above, the most influential attributes are size, industry, buyer type, requirements validation, and sales cycle movement. Have your Salesforce Admin add custom attributes on the Opportunity object for each of these factors, and assign a pick-list of value. For example, ‘size’ might have employee count ranges ‘<100’, ‘101-500’, ‘501-1000’, etc.

Okay, now back to the formula for ‘Win Probability’.

Assign a score 1-10 for each possible attribute value. If selling to any organization aside from small companies usually results in a loss, give ‘<100’ a value of 10 and the others a much lower relative score.

Weight each of the attributes on their relative importance. If you find that size is the single most influential factor and it’s doubly more important that anything else give it a .2x weighting. On the other hand, if validation is not so critical and you win plenty of deals without it in place give it a .05x weighting. Iterate on this process for each attribute.

The tricky part is working out the logic for the ‘Win Probability’ formula. What you end up with, however, is a percentage that corresponds to the probability for winning a deal at all, or, in other words, quantifying what your sales sweet spot looks like. The result? No more confusion. You’ll have a subjective view of Probability used for driving sales rep accountability AND an objective view based on the characteristics of your deals according to the data for delivering clear visibility up the chain of command.

3) Reports & Dashboards

Now that you’ve got the scoring model in place and you’ve separated the Probability into its distinct parts, you’ll need visibility to the new data. Add a report to called ‘Off Model Forecasted Deals’ and define it to be a listing of Sales Opportunities forecasted to close in the current month where there is significant divergence between the Forecast Probability and the Win Probability. This will tell you if your Forecast is overweight on hard-to-win deals, i.e., too much risk.

Share the report weekly with everyone, sales reps included. That way they will begin to see where/why the system sees gaps between their deals and the sweet spot. It is much easier to get a deal back on track when you know where the problem lies. Once your sales reps are on board deals will quickly begin trending ‘on-model.’

Great sales forecasting methods demonstrate you’ve got command over the details of your sales pipeline, and that’s a recipe for success when you’re trying to scale. Start upstream in the sales pipeline, add an opportunity scoring model, separate ‘win’ from ‘forecast’ probability, and set up a few reports. By making these simple changes, I promise you will see a reduction in slippage, an increase in win rates, and an over-all improvement in pipeline health within 90-days.

VP Marketing & Sales

Swayne Hill co-founded <a href="">Cloud9 Analytics</a> in 2007 and currently serves as its SVP Global Field Operations, focusing on building a world class sales and service organization as Cloud9 scales to meet growing demand. He also serves on the advisory board for Mintigo, an intelligent lead gen service leveraging Big Data and analytics, and is the author of the <a href="">Data-Driven Sales Management blog</a>. Currently, he is the VP Marketing & Sales at <a href="">Kaleo Software</a>.