Sujit: The same part can come from five vendors, or three, or however many. Each one carries its own brand name, and people prefer one over another, so I can see the reason behind having them. But as a supply chain professional, I would imagine that carrying five versions of the same product is a pain. Would you agree?

Planner: Yes. It’s a pain. You’ll see the same item number with a letter at the front or at the end, and that letter tells you whether it’s aftermarket or OEM, or more often which vendor it came from. Some of our vendors are the brand. Now we’re introducing our own brand on top of that. So it becomes very splintered. Instead of selling one item that’s more predictable, we have five that are all over the place in 75 locations.

Sujit: Then why carry five?

Planner: A few reasons. New vendors, or problems with a current vendor. We like at least two sources when we can, because there are real supply issues with car parts. Things happen in the world, and you can’t get product out of certain places. Then there’s pricing. A vendor raises their cost. We can’t sell at that price, so we look for another vendor, or we build our own brand to get a better price. And once those versions exist, customer preference makes it hard to take any of them away.

“

Sujit: And the demand on a lot of these is not large to begin with.

Planner: Right. A lot of our volume is low and very sporadic. Split that across five item numbers and 75 locations, and it becomes really hard to plan at all. The location matters more here than people assume. A customer isn’t walking in off the street. A shop has a car up on the racks. They need five parts, and if we don’t have all five at that branch, they aren’t going to buy two from us and three from a competitor. They’ll go to the competitor. We just lost the sale.

Sujit: So the split shows up at the counter, not only in the forecast.

Planner: That’s why we recommend not having five varieties. Have one, because it’s easier to forecast, you’re going to sell it, and we can make sure it’s in the right place.

Sujit: Who decides which versions go away?

Planner: Product management generally tells us which vendor brand they’re getting rid of or replacing. Our job is to look at what the demand says and push back where we need to. We’ll tell them, “Customers prefer that vendor, so I’m not taking it down as far as you asked, because they’re going to buy it.” If we have no forecast, we’ll have no product. Most of the time they go with our recommendation, unless there’s a reason they have to keep it.

Sujit: Once you agree that one version goes, what happens to its history?

Planner: If it’s systematic, the system knows which one replaces it and moves the demand across. We’ve been working on that with product management and inventory for about six months now. The rest is manual. We’ll take half the sales off one and put them on the other by hand.

Sujit: And this morning?

Planner: Three items. I went to product management and asked why we even have them. We aren’t selling them. Get rid of them. They agreed.

Sujit: Is there a version of this that doesn’t need a person to catch it every time?

Planner: On one product line, yes. Full units, the expensive ones, big enough that we can’t store them everywhere anyway. We built logic around the vendor code that says, “No matter what brand is selling, I only want to plan this one, and only in these locations.” The system pulls the history from every version under that code into one, puts it on the version we chose to plan, and places it where the product actually sells.

Sujit: And the customer who wants a different brand?

Planner: They can still order it. It becomes a special order. We just don’t want to plan five of them and hold that much inventory on product that expensive. That one has worked very well compared to anything we had before. I wish we could get there on more of the product lines.

Sujit: Does the same idea work for where the demand sits, not just which brand?

Planner: It does. We have stocking codes, and the logic says if something is stocked in a location, keep the history there and forecast it there. A product might start out in the DCs only, and as it starts selling somewhere else, we move the history to the branch closest to it. We had a high forecast in Texas where our DC is, and suddenly it was needed in New Mexico. We have a location there, so the history moved and we started planning it there.