Customer demand forecasting for distributors

Most distributors forecast products: how much wire, how many fittings, which water heaters to stock for winter. Very few forecast customers: when each account should place its next order, and how big it should be.

The second kind of forecast is the one that tells your reps who to call. And you can build a simple version from the order history you already have.

Why customer forecasting matters now

Large distributors are investing heavily in forecasting. Sonepar recently announced another $560 million in software to, among other things, forecast demand and optimize inventory, on top of nearly $4 billion committed to reshaping its business since 2021 (tED).

Independents don't need that budget to get the core benefit. Contractors already split their buying: HARDI's 2025 research found 83% of HVAC contractors use three or more suppliers. When one of them starts ordering later than usual, the first sign is in your own data. A customer forecast catches it.

What a customer demand forecast is

A customer demand forecast answers two questions for every account:

• When should they order next? Based on how often they normally order.
• How much should they order? Based on their normal order size, adjusted for the season.

When an account misses its forecast, that's a signal. Not proof that they've left, but a reason to look and probably a reason to call.

How to build a simple customer forecast

Step 1: Pull the right report

Export a sales-by-customer report with invoice dates and amounts for the last 24 months. Prophet 21, Eclipse and QuickBooks can all produce one. Two years matters, because it lets you see seasonal patterns twice.

Step 2: Find each customer's reorder cycle

For each of your top 100 accounts, calculate the average number of days between orders over the last 12 months. That's their reorder cycle.

A contractor with 40 orders in a year has a cycle of about 9 days. One with 12 orders has a cycle of about a month.

Also note how steady they are. A customer who orders every 9 days like clockwork is easier to forecast than one who orders anywhere from 3 to 20 days apart.

Step 3: Set the expected next order date

Add the reorder cycle to each customer's last order date. That's the date they should order next.

Step 4: Adjust for seasonal customers

Some customers don't follow an even rhythm. An HVAC contractor may stock up for heating season every fall and barely order in spring. For these accounts, compare this month with the same month last year and the year before, not with last month.

If a customer spent heavily every October for two years and hasn't this year, that's a missed forecast even if their regular cycle looks normal.

Step 5: Build the weekly overdue list

Every Monday, list the accounts that are past their expected order date by more than half a cycle. A 10-day customer at day 15, or a monthly customer at day 45. Sort by what they spent last year.

That's your reps' call list for the week.

How to tell a real signal from noise

Not every missed forecast means trouble. Check three things before calling:

1. Is the whole market slow? If most similar customers are behind too, it's probably the market or the weather.
2. Did something change on your side? A price increase, a backorder, a rep change or a credit hold around the time they slowed down is a likely cause.
3. Is it one product line or the whole account? A customer who still orders but stopped buying one category has moved that category somewhere else.

Common mistakes

• Using one cutoff for everyone. A 90-day no-order rule treats a weekly customer and a quarterly customer the same.
• Forecasting only totals. An account can stay on schedule while one product line disappears.
• Never updating. A customer's rhythm changes as their business changes. Recalculate at least every quarter.

Frequently asked questions

What is a reorder cycle? It's the average time between a customer's orders. A contractor who orders 26 times a year has a reorder cycle of about two weeks.

How much order history do I need to forecast customers? Twelve months is enough for a basic reorder cycle. Twenty-four months is better, because it shows seasonal patterns more than once.

Can a spreadsheet do customer demand forecasting? Yes, for a few dozen accounts. It gets hard to keep current for hundreds of customers, especially when you need to check seasonal patterns and individual product lines every week.