The Practitioner Files

Your Catalog Already Knows Where Demand Is Heading

One code on every part tells this team whether demand is fading or about to climb. A demand planning leader explains how they use it, and why they test the planning rules everyone else takes for granted.

A conversation between Sujit Singh, COO of Arkieva, and a demand planning leader at an automotive aftermarket distributor carrying more than a million SKUs across 75 locations.

Most planners find out where demand is heading from sales history. Our planner and her team found a faster signal, and it was already in their own catalog.

Sujit started by asking whether outside events, like a viral video or a news story, ever move her forecast. Almost none do. Tariffs, sometimes, and a competitor running out of stock. However, the signal that does move it’s coded onto every part her company sells.

What This Team Does Differently

WHY IT MATTERS: If history is the only thing telling you where demand is going, you’re always a cycle behind.

They read demand direction off the part.

Every part carries a code for the transmission speed of the vehicle it fits. A lower speed means an older vehicle, so demand for that part is fading. A higher speed means a newer vehicle that’s about to need repairs, so demand is about to climb. The team uses the code to keep the forecast from lagging in either direction, instead of waiting for sales history to catch up.

They test the rules they work from.

The company raised the forecast 20% when the lubricant line was promoted, and the team raised northern winter forecasts based on past snowstorms. Both looked reasonable. Both ran for two years. When the team checked them against what actually sold, neither had held up.

The Conversation, in Three Parts

Part 1. How the code on each part tells them where demand is heading

Planner reviewing digital charts with map

Sujit: Walk me through how a sunset or a new product shows up for you.

Planner: We’re on a monthly cycle like most demand planning groups. We build a forecast off recent history. But we also have coding on our product that tells us what speed transmission the vehicle takes. The lower the speed, the older the vehicle. The higher the speed, the newer it is.

Sujit: So the low speeds are on their way out.

Planner: Right. Those vehicles aren’t on the road much anymore, so we use the code to make sure we’re not over-forecasting. We’ll also go to our product groups and say, “We’re not selling this; so maybe we offer it as a special order instead of stocking it.”

Sujit: And the other end?

Planner: Ten-speeds, right now. Those vehicles have been on the road about eight years. The warranty’s out, and they’re starting to need repairs. Since we’ve been on Arkieva, this is the first time that we’ve been able to see what’s up and coming and filter on it.

“

This is the first time we’ve been able to see what’s up and coming.

Sujit: Does the forecast keep up?

Planner: I don’t know that any forecasting system could keep up with that growth. It grows really fast. So we do a lot of demand planning overrides. We look through the week at how it’s selling, whether the forecast was high enough, what our accuracy and bias were, and we adjust.

Part 2. What changed once someone’s job was to watch it

Her role didn’t exist before she arrived four and a half years ago. The signal was there, but nobody was responsible for watching it.

Sujit: What did this look like before your team existed?

Planner: I wasn’t here, but my understanding is the system didn’t keep up, and they didn’t have a team that looked at it. When something was up and coming, they’d just go across the board and add 10% to everything. Now we have a tool that lets us say, if we’re going to raise it, we’re going to do it right. The right product, the right locations, not everything just because it’s a certain speed.

Sujit: And the difference shows?

Planner: With the six-speeds, we were never ahead. With the ten-speeds, we’re keeping up with it more than we have in the past. And on the sunset product, the business can see we’re not ordering as much inventory.

Part 3. Checking the rules they work from

Two standing adjustments that looked reasonable, and what the team found when they checked them against actual sales.

Sujit: Do promotions move your numbers?

Planner: We don’t see a lot of increase from promotions. When I first started, I heard “We’re promoting our lubricant line, bump it up, we’re going to sell 20% more.” So we did. We’ve run the promotion for two years, and we’re not selling more. We’re just getting more inventory. So we come back with those results and suggest maybe we bump the safety stock in certain locations instead.

Sujit: Is there anything you look ahead to and say, when this happens, my demand moves?

Planner: Snow. Winter is when transmissions get used hardest. Plowing, pulling boats, pulling campers. But the past two or three years, the winters haven’t been as hard. We looked at history in certain northern areas for January, February, March, and we increased the forecast. Two years running, the weather wasn’t as bad, and we ended up way over forecast in those areas.

Sujit: And this year?

Planner: I’m in the process now of re-looking at all of it, because I’m hearing this one is supposed to be a worse winter. The question is, are we expecting snow, and enough of it that we’re going to see those repairs?

Sujit’s Read: What This Looks Like Across Companies

Most businesses have to infer where a product sits in its lifecycle from sales history. That means they find out late, on the way up and on the way down. What struck me about this team is that they don’t have to. One attribute on the part tells them which direction demand is heading.

The same goes for planning rules. A promotion uplift or a seasonal bump is a hypothesis. The teams I most respect are the ones that check that hypothesis, bring the result back to the business, and ask the question again before the next season.

TWO CAUTIONS

1. The signal gives you direction, not size. Her ramps still outrun the statistics, which is why her team overrides every week.

2. A signal only helps if someone is watching it. Before her team existed, nobody was looking.

A QUESTION TO TAKE BACK TO YOUR TEAM: Which attributes in your item master already tell you how old a product is or where it sits in its life, and is anyone watching it?

This is one piece in a longer set of conversations with practitioners about what it takes to stay ready for the disruptions that arrive in ordinary, unglamorous ways. More of these conversations are coming.