How to Choose the Right Forecasting Technique [+ Expert Insight and Data]

Forecasting can feel like a dark art — part science, part intuition, and a dash of hoping for the best. But as businesses face increasing pressure to predict everything from sales targets to inventory needs, relying on gut feelings just doesn’t cut it anymore.

I’ve spent weeks talking to forecasting experts, sales leaders, and business owners about how they actually approach forecasting (not just how they’re supposed to). What I discovered is that while the methods may sound intimidating, the core principles are more approachable than you might think.

Whether you’re trying to avoid another inventory stockout or looking to make smarter revenue predictions, I’ll walk you through the most practical forecasting methods and help you choose the right approach for your business.

Table of Contents

What Is a Forecasting Method?

A sales forecasting method is a systematic approach to understanding future possibilities based on both historical data and human insight. In B2B, this often means combining hard numbers (like pipeline data) with qualitative inputs (like sales rep confidence levels).

All of this can help you know what to expect the next month, quarter, and even fiscal year to look like.

“Forecasting feels like having a backstage pass to the future of our market,” says Chris Bajda, Managing Partner at Groomsday. “By tapping into data from previous seasons and current trends, we’re able to predict what our customers will need and when.”

I’ll share an example to help make this concrete. Imagine I run a coffee shop. A simple forecasting technique might just look at last year’s sales and add 10% for growth. But a more sophisticated approach would consider:

Seasonal patterns (iced drinks in summer, holiday drink specials).
Day-of-week trends (busy weekday mornings vs. leisurely weekends).
Local events (nearby office closures, construction projects).
Market changes (new competitor opening nearby).
Economic factors (inflation affecting coffee bean prices).

The impact of forecasting can be dramatic. An analysis by Dgtl Infra found that when they used integrated forecasts (combining sales data, usage metrics, and market trends), they closed 31% more revenue than those relying on pipeline data alone.

Source

Pro tip: If you’re looking to brush up on your forecasting skills, I recommend checking out these free courses in HubSpot Academy: Forecasting and Analytics in Sales Hub and Hubspot Sales Forecasting.

Types of Forecasting Methods

Forecasting methods generally fall into two main categories: qualitative and quantitative approaches. I like to think of them as the “art” and “science” of forecasting — both valuable, but used in different situations.

Source

Qualitative Forecasting Methods

Qualitative forecasting methods shine when historical data is limited or when you’re venturing into new territory. They rely on expert opinions, market insights, and informed judgment rather than pure numbers.

For example, if you’re launching an innovative product with no direct competitors, you might use:

Delphi Method (gathering expert opinions systematically).
Market Research (customer surveys, focus groups).
Expert Judgment (industry veteran insights).

Best for: New products, innovative industries, or sectors with limited historical data.

Quantitative Forecasting Methods

Quantitative forecasting is all about the numbers — using data-driven models to make predictions. Think of it as letting the data tell the story.

For example, a retail chain might analyze:

Past sales data across all locations.
Seasonal buying patterns.
Economic indicators.
Weather patterns.
Customer behavior metrics.

Examples of quantitative forecasting include:

Time Series Analysis.
Moving Average.
Exponential Smoothing.
ARIMA.
Regression Analysis.
Machine Learning Models.

Best for: Stable, data-rich industries where historical patterns can reliably inform future predictions.

TL;DR? Many successful businesses actually combine both qualitative and quantitative methods, using data to inform decisions while still leaving room for human insight and market knowledge.

Best Forecasting Methods

In speaking with dozens of experts for this piece, one thing became clear to me: There’s no consensus on what method is “best.” The options vary widely depending on your end goals, your industry, the data you have available, and much more. It will also greatly depend on which forecasting software you choose.

That being said, here are some top forecasting methods that you may find helpful.

1. Time Series Analysis

Time series analysis is widely used for recognizing trends and seasonality in historical data; it’s a heavy hitter in the forecasting world. Many experts that I spoke with use time series as one of their methods.

Bajda from Groomsday explains, “Time series analysis is especially useful for businesses that experience seasonal peaks and valleys, like retail.” This method helps track cyclical patterns, allowing businesses to optimize inventory and marketing strategies for anticipated demand changes.

Below I explain specific types of time series analysis.

Moving Average

This is like taking your business’s temperature over time — it smooths out short-term fluctuations to show the real trend.

Here’s a simple example:

Q1 Sales: $100,000

Q2 Sales: $120,000

Q3 Sales: $110,000

Q4 Forecast = ($100,000 + $120,000 + $110,000) / 3 = $110,000

Exponential Smoothing

Exponential smoothing is like your business’s short-term memory. Just as you would remember what happened last week more clearly than last year, this method gives more weight to recent events.

Here‘s a real-world scenario: Let’s say I run a downtown lunch spot. My sales might look like this:

Monday: $2,000

Tuesday: $2,200

Wednesday: $1,800 (Unexpected rain)

Thursday: $2,300

Friday: $2,500

A simple average would say I make $2,160 per day. But exponential smoothing might predict closer to $2,400 for next Monday because it:

Puts more emphasis on those strong Thursday/Friday numbers.
Considers the rainy Wednesday an outlier.
Spots the slight upward trend.

ARIMA Models

Auto Regressive Integrated Moving Average (ARIMA) is like having a master analyst who can spot complex patterns. While exponential smoothing is great for clear trends, ARIMA shines when things get messy.

Here‘s why it’s powerful. Let’s say I’m running an online fitness equipment store:

January starts strong (New Year’s resolutions).
Sales dip in February.
March sees a mini-surge (spring fitness push).
Summer is steady.
September spikes again (back-to-routine season).

ARIMA can handle all these patterns plus:

The lingering effects of past events (like how a viral TikTok video boosts sales for weeks).
Multiple seasonal patterns (daily, weekly, and annual cycles).
Irregular but predictable fluctuations.

2. Machine Learning Models

Machine learning has transformed forecasting by spotting complex patterns humans might miss. Dgtl Infra shared compelling results from combining AI with traditional methods.

Their data showed AI models identified enterprise user adoption growing 28% quarter-over-quarter, while sales team insights revealed financial services companies were integrating their API three times faster than other sectors — a critical pattern that pure data analysis missed.

They’re also the company I mentioned above that closed one-third more revenue when using an integrated forecast rather than just pipeline data alone.

Modern ML approaches include:

Neural networks: Identifying hidden patterns in customer behavior.
Random forests: Analyzing multiple variables like industry, company size, and usage patterns.
Gradient boosting: Improving prediction accuracy over time by learning from past forecasts.

3. Scenario Planning

In B2B, where single deals can make or break a quarter, scenario planning is essential. This method helps you prepare for different possible futures rather than betting on a single forecast.

“If we’re promoting a video for a seasonal campaign, like Black Friday, we create multiple outcome scenarios based on varying budget allocations, engagement levels, and ad placement strategies. This way, we’re prepared to pivot as needed,” explains Spencer Romenco, Chief Growth Strategist at Growth Spurt.

Here’s an example:

Conservative Case

— Only deals with 90%+ probability.

— Minimal upsell revenue.

— Standard churn rate.

Base Case

— Deals at 70%+ probability.

— Historical upsell rates.

— Normal market conditions.

Upside Case

— Additional stretch opportunities.

— Accelerated deal velocity.

— New product adoption.

4. Sentiment Analysis

Understanding the deeper context of customer feedback can be as valuable as tracking pipeline metrics. Sentiment analysis moves beyond basic satisfaction scores to uncover meaningful patterns in customer behavior and market direction.

For example, Kratom Earth incorporates feedback from customer reviews, social media comments, and direct interactions in their forecasting process.

“We pay attention to the words customers use, the benefits or effects they mention, and even any concerns they share. If we notice a trend where people talk about increased stress or a desire for relaxation, this guides us to forecast a higher demand [for certain products],” says Loris Petro, Marketing Strategy Lead at Kratom Earth.

“This allows us to plan inventory and marketing efforts around actual customer emotions and needs, which we believe is extremely accurate.”

How to Choose the Right Forecasting Technique

To illustrate how you can go through the decision-making process, I’m going to use a fictional example. We’ll call her Hannah and she runs an online pet goods store. Her orders have grown from 100 to 1,000 a month and now she’s facing some headwinds.

“I’m struggling to predict demand. Last month, I ran out of our bestselling cat food. The month before, I had to discount excess dog toys. There has to be a better way than just guessing!”

1. Take stock of your available data.

First ask yourself, what data do you have access to? Most businesses are sitting on more useful information than they realize. (P.S. This is where AI can be incredibly helpful!)

This could include:

Shopify sales history.
Purchase order records.
Customer reviews.
Email marketing metrics.
Social media engagement.

In Hannah’s assessment of the data, she might find that cat products make 45% of her revenue, dogs make up 40%, and other pets are 15%. In her business, she also sees seasonal trends that cause her products to spike — things like pet costumes around Halloween and new pet supplies around Christmas.

Pro tip: “If you have a strong history of data, methods like time series can reveal powerful patterns,” Badja suggests. For industries experiencing rapid shifts, machine learning models that continuously update based on new data are better suited to capturing real-time changes.

2. Connect trends from business patterns.

The next step is to go one step beyond the data — find ways to connect the dots.

In Hannah’s example, she might be asking herself:

“Why do certain products sell out while others sit on shelves?”
“How do holidays affect different product categories?”
“What’s causing these random spikes in certain items?”

By looking closely at the patterns over the past few months, you’ll likely spot some key trends. For instance, Hannah could discover that 90% of customers reorder every six weeks, sales spike after email promotions, and the weather doesn’t impact sales.

All of these discoveries offer helpful insight into her customer’s buying patterns and how she can better predict future sales.

3. Select your method.

Now comes the fun part — choosing your forecasting approach. Let‘s look at different methods through Hannah’s lens.

For example, if Hannah calculated the simple average across the last few months, she wouldn’t end up with any results that she could use to predict the future.

Simple Moving Average

Last 3 months sales:

January: 800 orders
February: 900 orders
March: 1,000 orders
Basic forecast: (800 + 900 + 1,000) / 3 = 900 orders

However, a multi-factor method could better account for her business’s growth rate and seasonal patterns.

Product Forecast =

(Base Average)

× (Growth Factor)

× (Seasonal Factor)

× (Marketing Impact)

Example for Premium Cat Food:

Base Average: 302 units

Growth Factor: 1.15

Seasonal Factor: 1.0 (non-seasonal)

Marketing Factor: 1.2 (email campaign planned)

June Forecast = 302 × 1.15 × 1.0 × 1.2 = 416 units

Pro tip: Make sure you are factoring in both qualitative and quantitative data.

4. Leverage short-term and long-term projections.

Start by mapping out sales projections for your specific business. Take a piece of paper and draw three columns: this month, this quarter, and this year.

For instance, if you run a software company, your immediate concern might be customer churn rate, while your quarterly view focuses on new feature launches, and your annual picture considers market expansion. A retail business might track daily inventory in the short term, seasonal trends quarterly, and store expansion annually.

Pro tip: “Don’t forecast based on past success,” says Stephen Do, Founder of UpPromote. You must consider uncertainty. Marketing changes constantly — new competitors, customer behavior, and affiliate marketing trends can disrupt your models.”

5. Build your integration system.

As I mentioned earlier, you’re likely sitting on a ton of valuable data — let’s put it to use.

To maximize forecasting accuracy, you can pair a CRM like HubSpot with an AI-driven platform, recommends Jeremy Schiff, CEO of Salesbot.io.

“While typical forecasting methods often focus solely on funnel performance, Salesbot.io leverages data across platforms like HubSpot to gain a comprehensive view of the entire sales pipeline — from lead generation to MQL, SQL, opportunity, and closed-won,” Schiff says.

“By aggregating insights from HubSpot, we can pinpoint which channels are working best at each stage of the sales journey, enabling smarter investment decisions and optimized resource allocation. This approach allows us to forecast not only future deal closures but also channel-specific effectiveness, helping us maximize impact across the sales process.”

6. Adjust on a regular basis.

This is where most forecasting efforts succeed or fail. You need a regular rhythm of reviews, but they should fit naturally into your existing workflow.

Forecasts aren’t one-size-fits-all. As Michael Benoit from ContractorBond says, “We review our forecasts every quarter to ensure they’re still relevant.” Regularly updating forecasts with current data helps businesses stay agile and maintain alignment with real-time conditions.

Pro tip: “When forecasting, especially with a team, you have to strike a balance between being too conservative and too ambitious,” Lexie Smith, Founder and CEO at Growth Mode, recommends. “Setting goals too conservatively may mean hitting targets sooner, but if they’re too achievable, it risks undershooting potential and can leave you vulnerable to unexpected shortfalls. On the flip side, overly ambitious targets can be unrealistic, leading to slow adjustments and missed opportunities for recalibration if early performance indicates underperformance.”

Improve Your Financial Health With Forecasting

After spending weeks learning from experts and business leaders about forecasting, here‘s what I’ve learned: Don’t get caught up in making things more complicated than they need to be. Your forecasting should actually solve real problems in your business.

Running a retail store and constantly running out of stock? Start by tracking your inventory patterns. Sales team missing their targets? Focus on those pipeline metrics.

One thing that really stuck with me was that buyers rarely follow a perfect, linear path. Your forecasting needs to roll with the punches when your assumptions turn out wrong.

Sure, we‘ve got more powerful forecasting tools than ever before, from basic spreadsheets to fancy AI systems. But at its heart, good forecasting isn’t rocket science: get reliable data, find patterns that actually mean something, make smart predictions, and learn from what really happens.

My best advice? Start with whatever matters most to your business right now. You can always build from there.