Types of forecasting: which is best for you?

Top-Down vs. Bottom-Up Forecasting for CPG Brands

Developing reliable sales forecasts is a crucial capability for CPG brands of all sizes. Accurate forecasts enable smarter budgeting, production planning, and strategy development.

But how should brands build their forecasts - using a top-down or bottom-up approach? Here we’ll explore forecasting techniques, how forecasting techniques work, weigh up the pros and cons, and look at when each one is most appropriate.

What is Top-Down Forecasting?

A top-down forecast is a forecasting technique that begins at the highest level of a business’s sales structure and divides targets across lower levels from there.

Here’s what it looks like:

1. Start with an overall revenue target for the coming year based on leadership goals. For example, achieving $100 million in sales. 

2. Break this total target down by region. For instance, attributing $50 million to the West Coast and $50 million to the Midwest. 

3. Drill further down to allocate revenue targets to major customers in each region. Such as projecting $25 million each from retailers A, B, and C in the West.

4. Continue to the lowest level, assigning sales targets for specific products for each customer. Like $5 million of Product X to Retailer A.

The process flows top-down, taking a high-level target and layering on assumptions to distribute the numbers down the organization structure.

Pros and Cons of the Top-Down Forecasting Technique


Quick and straightforward to implement - a top-down approach can be put in place rapidly without complex analysis. This enables faster planning.

Directly aligns forecast with management expectations and business goals - top-down numbers originate from leadership targets. This ensures the forecast matches desired outcomes.

Requires minimal historical data analysis - tithout analyzing granular historical data, assumptions are made quickly. This is helpful when historical data is lacking.

Provides a clear strategic roadmap - with targets set across regions, products, and accounts, the forecast supplies a simplified path to achieving goals.


Less accurate since assumptions may be unrealistic - without a grounding in historical data, unrealistic assumptions can creep in, which can undermine accuracy.

Hard to explain or defend specific numbers - since targets lack clear analytic rationale, they can be hard to explain to stakeholders.

Can discourage collaboration and buy-in across the organization - numbers handed down from above feel detached from reality for lower-level teams, which can hamper buy-in.

A top-down approach works best when:

  • Senior leaders have a clear growth target or strategic vision.
  • Limited historical data is available.
  • Business is rapidly evolving or entering new markets.
  • Simplicity and alignment are priorities.
What is Bottom-Up Forecasting?

In contrast, bottom-up forecasting uses a granular view of historical data to build projections from the ground up.

Here’s how it works:

1. First, you’ll gather highly detailed sales data at the lowest level - individual products or SKUs.

2. You’ll then use statistical models to project sales trends for each product based on its historical performance.

3. After that, you’ll roll up the individual product forecasts to summarize at upper levels like customer accounts, regions, and total company, and collaborate with managers at each level to refine and add judgment to the projections.

4. Finally, you’ll arrive at an overall forecast that aggregates the many bottom layers.

Pros and Cons of the Bottom-Up Forecasting Technique


Increased accuracy based on granular historical data - aith data like sales by product/account, trends are projected accurately from historical performance.

Easily explainable since each number has a clear rationale - the data behind each forecast number enables logical explanations when questioned.

Collaborative process fosters organization-wide engagement - collecting bottom-up data is an inclusive process that gets broader team input.

Uncovers growth opportunities at lower levels - granular data highlights standout products/accounts to focus innovation and expansion efforts.


Can be data-intensive and time-consuming - extensive data collection and crunching to model product/account trends takes major time.

Requires large volumes of clean, granular historical data - without sufficient transaction-level history, modeling, and extrapolating trends is impossible.

Hard to align fully with subjective executive targets - data-driven forecasts may clash with management expectations, requiring reconciliation.

Bottoms-up forecasting works best when:

  • Rich historical data exists for analysis.
  • Business is stable and seasonal patterns are predictable.
  • Accuracy is highly valued.
  • Consensus and coordination across the organization is important.

Blending Forecasting Techniques for Maximum Impact

While each method has pros and cons, combining top-down and bottom-up forecasting often yields the best results. 

Top-down forecasting provides the big-picture view and clarity on corporate goals that ground the process. Bottom-up adds nuance, accuracy, and participation from the full organization. 

Using both together, leadership sets the strategic vision while revenue managers fill in the details. The result is an accurate, organization-wide forecast that balances vision and precision.

Forecasting with Confidence

Forecasting is both an art and a science. The right data, models, and processes can steer brands toward forecasting success. But effectively blending analytical rigor with experience and judgment is what truly unlocks precise projections.


Forecasting Trade Promotions vs. Forecasting Sales for CPG Brands

Own Your Future With Intelligent Forecasting.

Forecasting in Excel is tedious, time-consuming, and error-prone.

Take the guesswork out of forecasting. Vividly automates the entire sales forecasting process and updates any changes in real-time.

See how industry leaders like Liquid Death and Perfect Snacks benefit from Vividly’s intelligent forecasting.