Sales forecasting is a necessary — occasionally painful — part of preparing for the upcoming fiscal year and managing sales goals along the way. Since leaders can’t use a crystal ball to predict the future, they are left analyzing quantitative, and sometimes qualitative, data to anticipate future sales.
But this sales forecasting process becomes problematic when sales teams and executives confuse “optimistic goals” with “accurate forecasting.”
Instead of looking at historical data and making forecasts based on previous trends and realistic parameters, salespeople (who are optimistic by nature) tend to create forecast numbers weighted toward the best hopes of the sales team and C-suite.
In this article, we’ll break down the best strategies and tactics you can use in your next sales forecast to better predict your sales team’s success next year.
Ironically, excessive optimism in the sales forecast often creates unnecessary negativity and disappointment among team members down the road. It’s better to identify and exceed realistic targets based on solid data than it is to set the bar unreasonably high and miss the target.
We’ve talked so far about the importance of sales forecasting, but how do you actually do it, and do it well? The best sales leaders use something called a forecasting model.
There is no one-size-fits-all forecasting model, though. When it comes to high-performing sales teams, you’ll need to choose the right model (or combination of models) to create a fair and accurate annual goal. Check out some of these popular forecasting models below. While all of them can’t be used for sales on their own, you use the information you gather from them to help inform your sales forecast.
1. Length of Sales Cycle Forecasting Model
An important factor for every sales rep (no matter their industry) is the length of the sales cycle. This is essentially the time it takes for a prospect to pass through every stage in the sales cycle — from lead all the way to repeat buyer.
Understanding the length of your sales cycle and factoring that into your sales forecast will help your sales team focus on closing deals rather than rushing prospects through the process in order to beat the clock each month.
Consider the length of your sales cycle as a fixed metric. Unless your sales leadership team has plans in place to speed the cycle up, go ahead and assume that the length of the sales cycle won’t change — and don’t encourage your sales reps to rush their deals either.
2. Time Series Forecasting Model
If you can’t tell by now, using historical data to forecast for the future is the theme in many of these forecasting models, but time series forecasting focuses primarily on historical data without the use of other variables. With this type of forecasting model, your sales will be plotted on a line graph with each point representing a specific point in time.
You can use time series forecasting to predict when future sales might happen based on when sales have happened in the past.
3. Demand Forecasting Model
There are a few different types of demand forecasting models that focus on internal and external factors that affect demand. To keep things simple, we’ll focus on the two most popular types of demand forecasting: passive and active.
Passive demand forecasting looks at past data to predict future sales. This type of model keeps things simple by only accounting for internal factors that your business can control. Seasonal trends that your business typically experiences are also taken into account.
Active demand forecasting uses data in real-time (or as close to it as you can get) as possible to forecast future sales. With this type of model, you’ll include external factors like the state of the market, marketing strategies that are currently in play, and knowledge of the competition if you have it.
If you’re looking for a comprehensive guide on demand forecasting models, I recommend you take a look at Jake Rheude’s article over at Red Stag.
4. Regression Forecasting Model
It’s time to roll up your sleeves and get hands-on in Excel for this model. A regression model is a statistical process for understanding what independent variables are affecting your dependent variable. In sales terms, a regression model helps you understand which sales activities affect how many deals you close.
The formula for a regression model is Y = a +bX, where Y is the dependent variable and X is the independent variable. Values a and b are the y-intercept and slope of the regression line, respectively — don’t worry, Excel will take care of a and b in the formula automatically when you supply historical data about your sales.
The key to running a helpful regression forecasting model is using historical data and running a regression for each activity. Once you’ve run your regression model, you’ll be able to see a correlation (or lack thereof) between a specific activity like “emails sent” and your number of closed won deals over time.
Some additional activities that can affect your sales include the number of outbound calls made, inbound calls received, and demos completed. You can run a regression for each of these independent variables to determine which activities to focus on and which don’t matter much in the grand scheme of things.
Pro Tip: Remember, statistics do not define causation, only correlation. That means you should supplement the information you glean from the regression forecast model with qualitative data from your sales reps about what activities are the best uses of their time.
Most businesses experience some seasonality in their sales so it’s a great practice to account for this in your sales forecasting. A seasonal forecasting model can reveal exactly how much deviation each month has compared to the annual average.
To use data from a seasonal forecasting model, you’ll need to compare the seasonal index, a comparison between a specific seasonal time period to the average seasonal time period, to the average annual sales cycle.
For a detailed tutorial on how to use a seasonal forecasting model, take a look at this video.
Let’s look at a few simple tactics that sales teams and executives can use to create better forecasting models for their business.
1. Use historical data.
Most large companies have historical data they can use to determine realistic sales forecasts. If your company hasn’t implemented analytics and other forms of tracking methods that can be tied to goals and conversion rates, do so now. You need to know where you’ve been so you can accurately forecast where you’re going.
It’s true past sales are not always accurate predictors of future performance. This year you might release new products, expand into new markets, face an increase in competition, and so on and so forth.
But historical data is a solid foundation on which you can stand as you weigh additional, unpredictable factors that could increase or decrease sales in the upcoming year. These are scenarios you can weave into your presentation of firm numbers for your final forecast.
2. Keep clean records.
If no clear standards are communicated to the team, sales reps may come up with their own definitions and use cases, leading to inconsistent data entry. Or, if they don’t know how important a property is, reps may fail to use it altogether.
You can’t make good decisions on dirty data, so for any numbers that aren’t as concrete as sales and revenue — like current deals in the pipeline or number of deals per customer segment — make sure your team is on the same page.
You can do this by:
- Providing ongoing training to the team on CRM use
- Continuously referring to the forecast in team meetings
- Checking up on deals during one-on-one meetings
- Performing spot checks on records and deals to note inconsistencies
3. Start with a simple model.
I know it’s tempting to try and incorporate each of the model types we talked about earlier but resist the urge to do this. If this is your first time using a quantitative forecasting model to predict sales for the next year, don’t be afraid to start small and improve your model over time.
Using something simple like a regression forecasting model for five of the most common sales activities your team performs is a better model than one that combines seasonality, time series, and demand forecasting into one. Why? Because the fewer variables you have to keep track of, the simpler it will be to:
- Achieve your sales goals,
- explain to your sales reps how why the goals were set this way, and
- get approval from leadership on your forecast.
Once you determine how well your forecast model is working for the first year, you can update it the following year with variables from another type of model.
4. Implement a sales pipeline action plan.
For sales leads, quality is more important than quantity. While a lead’s quality can certainly affect its conversion potential, an increased quantity of leads typically increases the number of closed deals.
That’s why you should build an action plan for generating the minimum number of leads necessary. For example, if you know your reps close 25% of their deals from well-qualified leads, you may aim to generate twice as many well-qualified leads next quarter. Ideally, your reps will close 30-50% more deals.
No matter what your numbers need to look like on the closing side, put the same level of focus in forecasting and generating leads. Understand your conversion rates at each stage of your sales funnel, then plan accordingly.
For example, ask your sales team:
“What does it take to move a prospect through your sales process from the first inquiry to the final deal closing?”
“How many steps are there in your sales process, and what percentage of your leads (approximately) convert at each step of the process?”
“What is the definition of a ‘well-qualified’ lead? Is it someone who has gone through an online demo, someone who has filled out an intake questionnaire…?”
“Based on the conversion rates at each stage of your sales process, how many leads do you need to generate in order to achieve an expected number of sales?”
Pro Tip: Do the math by working backward through your sales process. For example, if you want to close 100 deals this year, and your salespeople close 10% of deals with leads who have already watched an online demo of your solution, and 10% of new inbound sales leads agree to sign up for an online demo, you need to generate 10,000 new inbound sales leads to make 100 sales: 10,000 x 10% x 10% = 100 sales.
The conversion rates and correct numbers for your pipeline will differ depending on your business and average deal velocity. This information lets you build an accurate sales forecast based on stage-by-stage conversion rates.
5. Use forecasting tools.
You can save a lot of time (and improve the accuracy of your forecast) by using a tool developed just for forecasting.
For instance, HubSpot’s sales reporting tools include a forecasting report that relies on your sales pipeline plan combined with current deal data to provide a forecast based on closing probability.
6. Incorporate “what ifs” and qualitative data.
Many companies fail to plan for new sets of data to track and overlook qualitative data. Instead of constantly looking at the same numbers and making bold predictions, companies should ask “what if” questions that can be answered once more data is collected.
Looking at your business from different angles gives you new insights. For instance, if you are trying to boost sales for multiple products on your eCommerce site, why not track how many customers purchase a top-selling product from two different categories? Understanding where customers gravitate to for certain items and which items pair well together could give you inspiration for new product promotions and special offers.
Qualitative questions paired with quantitative tracking can help you better understand your business and make smarter decisions. This is how you can integrate forecasting into other business objectives, such as remodeling a store or testing advertising campaigns.
7. Consider seasonality as a factor.
One type of qualitative piece of information is the answer to this question: “We sell more when…”
If your forecast is linear, treating every month and quarter similarly, you may lose accuracy on account of seasonality or related factors.
Here are a few examples to demonstrate this idea:
“We’re a toy company, and our sales go nuts around Christmas.”
This company would consider increasing the forecast in Q4, especially after Thanksgiving leading up to Christmas.
“We sell office equipment to office managers. That means we sell more during the business week when they are on the clock.”
If this company has a month with a lot of holidays (e.g. December), they should factor this in as a lower sales month in the forecast since office managers will not be in the office making purchases. In addition, they should also consider how the months fall and make accommodations for months that have fewer business days than others (e.g. February).
“We’re a roofing company, and we sell best when our customer is experiencing a roof leak.”
Even though roof leaks don’t have a seasonality, this company’s customer may not realize they have a roof issue until they see physical evidence of it (a leak). That means rainy seasons could result in more business, and they should consider factoring that into their forecast.
8. Encourage collaboration between all departments.
A well-constructed forecast often isn’t the byproduct of any single department’s contribution — it tends to incorporate input from across the company. Collaboration offers a new perspective to a company’s forecasting process.
Forecasting works best as a team effort. Incorporate input from multiple — if not all — departments at your company. Different departments have their own expertise to offer, allowing you to have a more well-rounded forecasting process.
Those contributions will also add a new degree of accountability to your forecasting efforts. If your process is rooted in teamwork and subject to more scrutiny, no individual department will have the space to adjust data to suit its interests and biases.
Additionally, inter-departmental collaboration adds an element of trust to your forecasting process by including diverse perspectives and helping departments feel heard.
9. Incorporate external data where appropriate.
The default when sales forecasting is to rely on internal, historical data that’s easily accessible. While this is an important piece of the puzzle, you’ll be able to create more realistic forecasts if you incorporate external factors into your model.
Is your marketing team running a big campaign next year? Did a competitor recently change their product or service? Has the market your serve expanded or contracted? Each of these external factors will have an effect on your business, and consequently your sales.
10. Consider market trends and competition.
Wouldn’t it be awesome if the variables that affected sales were all internal, such as sales team head count and effectiveness? However, there’s a whole host of variables related to market trends that affect sales.
Let’s say you have one product that is a steady staple and another that’s new, trendy, and receiving a lot of buzz but hasn’t caught on mainstream yet. These two products would not have the same growth trajectory, so it’s important to factor them in as separate segments.
Another thing to consider is competition. Let’s say you have a competitor with the same authority and awareness in the market as your organization. Their offerings are competitive, and they’re a great company. Then, they lower their price.
Something as simple as this changes the conversations reps have with prospects… and the conversations prospects have with themselves.
Continuing to keep a pulse on what the market is doing will help you create more accurate predictions.
11. Hope for the best, and prepare for the worst.
Few people enjoy thinking about worst-case scenarios, whether you’re talking sales forecasts or sports predictions. As much as you want to hit massive sales numbers every quarter, you also know the chances of your favorite team winning the Super Bowl.
That’s why our sales forecasts should always consider the worst that could happen: What if you lose your top three reps to a competitor, the product you’re selling faces an embarrassing recall, or something goes wrong that forces you to re-evaluate your sales process? You don’t have to spend too much time dreaming up the most horrific events your company could face, but you need to leave some cushion in your forecast that accounts for potential setbacks.
Scrutinize last year’s numbers — what went exceptionally “right” last year that might not happen again? What strokes of good luck did you have that might have made your numbers look better than reality?
Don’t assume every bit of good fortune is going to happen for you every year. The reality of sales numbers often lands somewhere between “the sky’s the limit” and “the sky is falling.”
12. Refer to Your Forecasts Consistently.
It might go without saying, but your forecasts are essentially useless if you don’t use them as reference points, so be sure to refer to them on a consistent basis. They’re crucial resources for guiding a wide variety of business decisions, including budgeting and directing marketing efforts.
Your forecast is never going to pan out exactly as you’ll plan. There’s bound to be some give and take between your projections and your actual figures. Still, you need it to have some concept of what the future might look like, a perspective on whether your performance is in line with your goals and expectations, and whether you’re allocating your resources effectively. Constantly keeping tabs on your forecasts is one way to ensure you’re covering all those bases.
Keep in mind that sales forecasting is not a one-time “start of the new year” activity — it’s an ongoing process that affects every aspect of your sales pipeline. And sales forecasts are not set in stone. They are “living documents” that help the sales team stay on target throughout each quarter.
With a data-guided process and plenty of open communication and collaboration, you can create more accurate sales forecasts and maximize your sales team’s potential.
Editor’s note: This post was originally published in June 2020 and has been updated for comprehensiveness.