for CPG Brands
UNDERSTANDING FORECASTING
What is forecasting?
Whether you're just starting out or are a more mature brand, having reliable forecasts is vital for steering your business toward growth and profits.
But before we can dive into forecasting best practices, we need to understand what forecasting is and why it’s so important.
At a basic level, trade promotion forecasting is simply the process of predicting your future sales (called Sales Forecasting) or your future demand (called Demand Forecasting). It influences estimating upcoming demand for your products across customers, regions, channels, etc. This enables you to see what your revenue, costs, and other key metrics are expected to look like in the future.
Forecasting typically examines different periods - next month, next quarter, next year. Most brands do annual, quarterly, and monthly forecasts.
The goal is to predict performance at a granular level across:
Different Products/SKUs
Different Products/SKUs
Various customers and channels
Baseline vs. incremental
This level of detail allows you to paint a comprehensive picture of where your business is headed.
We know predicting the future sounds about as easy as knitting a sweater for a snake. But here's why locking down accurate forecasts is so crucial:
Easier Budgeting
With expected revenue and sales targets in hand, you can smartly allocate trade promotion dollars where they're needed most. No more guessing how much to budget!
Informed Strategies
Forecasts highlight growth opportunities to double down on and risk areas to navigate, which can guide your marketing, innovation, and expansion strategies accordingly.
Optimized Operations
Hitting production targets hinges on forecast accuracy. Nail your forecasts to supply the right inventory at the right time.
Proactive Planning
Seeing potential sales shortfalls or spikes early allows you to adjust plans before it's too late. Forewarned is forearmed!
Getting your forecasts wrong means you also get trade spend, inventory, revenue targets, and ultimately, profits wrong. That's why this isn't an area to wing or settle for "close enough!"
This is always the first question to ask. Where does the data come from? What does it look like?
Stores keep a record of their sales in some way. Then they will furnish this data for other people’s use. Most retailers share their data, known as point-of-sale (POS) data, with third-party companies such as SPINS, Nielsen, or IRI, which compile data from many sources and produce reports. Some retailers like Whole Foods provide their own reports.
For a CPG company to get its own sales data, they must find the POS data by going to the source. They will visit either the store’s website, the SPINS website, or sometimes the distributor’s portal (for example, UNFI) for this information.
There, the data will be broken down by product, region, time period, or any other category. A CPG company should be able to download only the data that is relevant for themselves.
While the format of the data may vary between any two sources, there are some common pieces of information we care about for a given product, customer and time period:
Sales — the number of units sold. This is usually broken down by week.
Revenue — the amount of money from selling these units.
Stores — the number of stores that sold the product.
The following pieces of information do not necessarily come from a data source, but can be gotten by using an advanced trade promotion management and trade promotion forecasting platform.
Velocity — roughly speaking, sales divided by stores; a measure of how well the product sells in a vacuum. This is a derived value, calculated from other values and is typically divided by a certain timeframe, such as weekly or monthly. A related term is distribution, which is how widely your product is sold.
Seasonality — a multiplier of how well or poorly the product sells due to seasonal variations. The base sales number is to be multiplied by the seasonality; for example, if seasonality is 0.8, it means the product sells only 80% as well as usual during this time period, and if seasonality is 1.1, it means the product sells 110% as well. A value of 1.0 means business as usual.
Lift — a measure of how much sales were elevated during this time period. For example, if lift is 0.5, it means that 50% more units than usual were sold, and a value of 0 means no change. Lift is the result of trade promotions; when we run a trade promotion we expect (or hope) that the lift will reach a target amount.
Promotion status — whether or not a promotion was in effect during this time period.
Data Complexity
With multiple data sources to manage, keeping track of everything in tools like Excel can be overwhelming. This complexity often leads to errors, especially when data needs to be updated or refreshed frequently.
Lack of Real-time Updates
In a fast-paced market, real-time data is crucial. Traditional methods, reliant on manual updates, often struggle to keep pace, leading to decisions based on outdated information.
Addressing Data Gaps
Not all retailers and sales points have accessible data, especially smaller, local outlets. This lack of comprehensive data can result in skewed forecasts.
As technology advances, promotional forecasting tools are evolving. When you're evaluating tools to help you forecast, look for modern solutions that can help you focus on the following.
Automation
The aim is to reduce the manual workload, enhance accuracy, and allow forecasters to focus on strategy and interpretation rather than data entry.
Data Synchronization
Integrating multiple data sources into a unified platform reduces complexity and ensures all teams access the same, up-to-date information.
Scalability
As brands grow, their forecasting needs become more intricate. Modern tools are designed to scale with businesses, accommodating increasing complexity without compromising accuracy.
Versioning
Given the dynamic nature of the market, it's beneficial to have multiple forecast versions based on different scenarios. This flexibility allows brands to be prepared for various outcomes.
By now you're probably itching to perfect your forecasting capabilities. So what does the sausage-making process actually look like?
Here are key steps brands can take:
1. Start with the past - Analyze at least 12 months of historical performance data across products, accounts etc. Last year's sales give context for modeling the next.
2. Layer in assumptions - Factor in expected growth rates, seasonality patterns, product launches, account additions etc. to shape the forecast narrative.
3. Build statistical models - Create algorithms that crunch numbers to predict most likely outcomes based on past trends.
4. Add judgment - Sprinkle in expert human judgment to account for outside forces like economic conditions, competition moves, or gut instincts.
5. Create forecast scenarios - Build alternate optimistic and pessimistic projections to stress test different versions of the future.
6. Track accuracy - Compare actual results vs. forecasts to monitor predictability and quickly adjust course as needed.
It takes a combination of data analytics, business acumen, and course corrections to nail forecasts consistently.
.svg)
NEXT UP
Types of Forecasting and which is best for you
.svg)
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.
