Sales Forecasting – Where do I start?
In a previous article, I started reviewing some of the reasons why every business should have a better handle on their ability to forecast incoming sales, as accurately as possible. This is obviously easier said than done, especially since there are so many different ways of slicing the information.
And if the process wasn’t complicated enough, each CRM vendor pushes their own set of beliefs and methodology on their users.
The good news: you are not locked in a specific methodology. In this article, we will start reviewing some of the ways that a business can start forecasting their sales.
Wait, do I need anything to start applying forecasting models?
The most important thing to have obviously is a good understanding of your Sales process, which means it has to be documented (stages, activities and expectations), clear to everyone (i.e. communicated) and followed (ideally implemented with a CRM, able to track progress through it). I will review this part in more details in a following post.
To accurately forecast sales in a given period, all activities, progress made, quality of the relationship, criterias for the buyer to make a purchase, etc.. would be thoroughly analyzed and monitored constantly to quickly identify any opportunity that is slipping out. This is easy to do when the number of deals that are worked in parallel is limited. In today’s reality though, and as the organization scales up, this becomes extremely time-consuming and not necessarily the most productive.
So, what are the common types of sales forecasting out there?
The first type of “forecasting”, and probably the most traditional way of doing it, is to only consider the sales rep (or sales manager)’s feedback in terms of the chance of closing. Organizations using this model will rely on feelings and emotions of the sales reps, based on the last interactions with clients to quantify how many opportunities will convert and when. Sales reps will be asked to commit on a number, with little regard to the entirety of their pipeline, or any progress. Sales managers will then usually apply a percentage on the sales reps numbers, mostly based on historical trends and the tendency of that specific sales rep to either overachieve or overestimate the number they commit. We will consider this methodology as the empirical forecasting model.
Typically, if a CRM is in place, the most simple way to forecast sales is to track progress against the different stages, assign a probability of closing to each stage (typically around 5-10% at the discovery stage, 50% at demo, 75% at proposal and so on). Managers can then derive a weighted value of all opportunities in the pipeline set to close in a specific period, by simply multiplying each opportunity amount by its probability of closing and then adding them all up. This is what we could consider as the typical progress-based forecast.
A third option, still leveraging CRMs, is to start categorizing opportunities by a number of criteria, and then assigning weights on each category to roll up the overall company forecast. Most commonly used categories will be buckets that indicate the current chance of closing, such as pipeline, best case and commit. This category-based forecast brings a layer of sophistication to the typical progress-based forecast, by suggesting that opportunities that may progress at the same pace don’t necessarily have the same chance of closing due to other criterias (prime vertical or not, relationship with contacts within, etc..).
Which one is the right one for my business?
As one can imagine, there is not a one-size fits all solution to everyone’s different business model. Some forecasting strategies are better adapted to low-volume high-complexity sales model, while they will be considered overkill in a high-volume low-complexity sales model. The best is to review what makes the most sense to your organization, and usually, it will include a slight redesign of the sales process to ensure proper alignment with the forecasting model.
With the arrival of Artificial Intelligence (AI) in CRMs, it opens up the door for additional options, but even without the implementation of complex systems, there are a handful of things that can be put in place rapidly to increase the sales forecasting accuracy.
Interested to talk about forecasting models in more details? Give us a call (or fill the form below) and we will be happy to review your model together.
[hubspot type=form portal=5738168 id=e652fa22-6774-4a2d-9522-3ad4fd009b15]