It pays to be right about the future - which is why a lot of analysts spend a lot of time creating forecasts.

But when you sit down to actually make a forecast, it can be like staring at a blank sheet of paper - you just don’t know where to start.

So let’s break it down into steps.

Step 1: Create the base forecast

Suppose you are trying to create a yearly sales forecast for your product (or service). You need a starting point. Where you start depends on whether you’re dealing with a completely new product or an existing one.

If it is a completely new product:

Start with market research to estimate demand and potential uptake. Market Research is vital to estimate both the potential market size as well as the rate at which you can grow to that size.

Even a small, well designed consumer survey can yield valuable information about potential customer uptake.

And desktop research into similar products in the market, or the same product in other markets can get you some good starting numbers for sales and growth

All this will help you create a base forecast. However, forecasts for new products are more likely to be off, so it's essential to build scenarios (explained in Step 3).

If you have previous sales data:

Use time series analysis to get started. Look for trends, seasonal patterns, and other cycles in your historical data that can be projected forward.

Seasonal decomposition is an excellent technique to break down historical data. Many analytic tools offer this feature, or you can use Python or Excel if you're familiar with them.

Once you understand the sales trends and seasonal cycles, project forward, assuming similar patterns to begin with.

But even with sales history, validate your forecast by looking at other markets and research to ensure the trends will likely continue.

Step 2: Fine-tune with Externalities

Next, adjust your forecast for external factors—variables outside your control that affect sales. While there can be many, always consider these three key areas:

1. Economic Forecast:

• Start with a simple GDP growth forecast (available from financial institutions). Add inflation or currency devaluation if relevant.
• Remember: Most products grow in a growing economy, but some products can thrive in a slowing economy (e.g., discount retailers), so the GDP’s impact requires informed judgment.
• Based on these you might increase or decrease the growth rates in your forecast

2. Competition:

• Competition can influence growth in unexpected ways.
  • Early-stage markets often grow faster with more competitors because of increased awareness and adoption.
  • In mature markets, aggressive new entries can disrupt pricing and demand with oversupply.
• Adjust your forecast based on the competitive landscape.

3. S-Curve Dynamics:

• Growth rates typically follow an "S-curve" as industries mature:

• Early-stage growth is slow but accelerates dramatically once the market reaches a tipping point.

• Mature markets often see growth rates slow significantly.

  Positioning your product on the S-curve can be challenging, as S-curves vary widely across industries. If you're unsure about where your product stands, it's better to be cautious and avoid relying too heavily on this model.

  However, when market indicators are clearer—supported by deep research or comparisons with similar markets—S-curve insights can be invaluable for refining long-term forecasts. Use them as a guide, but stay adaptable to new information.

Step 3: Apply Judgment / build scenarios

By Step 2, you’ve already begun applying judgment to account for externalities. But creating a forecast often requires collective wisdom—market knowledge, internal insights, and awareness of upcoming product features.

Since judgments can vary widely (and be wrong), mitigate risks by building multiple scenarios:

• Create a pessimistic, optimistic, and base case.
• These scenarios allow you to see the impact of different judgments and associated assumptions on your forecast and help align your strategy with your risk appetite.

The beauty of scenarios is that they turn vague judgments into tangible numbers, giving you actionable options to choose from.

Step 4: Monitor & Update

No forecast survives impact with reality unscathed. It will always be at least a little off—sometimes massively so.

Think of a forecast as a living document. As events unfold, validate, correct, and recalibrate it to improve accuracy.

• Don’t cling to your original estimate or be defensive about adjustments.
• Use new data, market shifts, and competitor actions to refine your forecast continuously.

As time moves forward, historical data grows, new externalities emerge, and the forecasting cycle begins again.