Thursday, August 19, 2010

Supply Chain Juggling Act: RIM Programs, Facing Fill, Network Density, and :-)”

One of my favorite industry friends is brilliant and very polite. She sent me an email challenging my discontent with facing fill as a key strategic metric to manage, especially under Retail Inventory Management (RIM) environments.
“After our brief discussion of the importance of Facing Fill in a RIM world, I thought I'd send you some more information. From the dealer's point of view, since they do not need quick ORT for RIM recommended parts (assuming that they allow RIM to manage them), the decision to stock the part in the facing PDC becomes a cost trade-off decision. However, for the parts that RIM does not recommend, facing fill has the same importance as before RIM - it continues to be the biggest driver of dealer satisfaction. (See - I have been listening to you all these years). …
For any one dealer, RIM does a great job at recommending the stocking strategy. But, when you try to look at the part level, very few parts are recommended to most of the dealers. I know that this is a rather simplistic look, but it gets the point across. Until we find a way to have more of the total volume of a part RIM recommended, it will still be important to stock the part close to the dealer.”
She closed the email with:
“Please feel free to correct my way of thinking - I know that I can count on you to point out the flaw in my logic. :-)”
I love those smiling faces you can make with punctuation. They kind of humanize an email.

Her logic is simple about RIM-managed parts. If you cannot get a significant portion of your dealers using RIM for each part number, then facing fill is critically important and translates into dealer satisfaction and end-customer satisfaction. You could even diminish dealer purchase loyalty by having low facing fill. Because of all this, facing fill is a very important metric.

OK, yes and no.

First let’s talk about why the answer is “yes.”

Facing fill measures order completeness within a typical order cycle (order response time – ORT) from the “facing” warehouse - the warehouse assigned to each dealer for their stock orders. If your warehouse network has been collapsed to support a RIM strategy (fewer facing warehouses), ORTs for non-RIM orders will be longer. Longer ORTs could result in dealer defections for parts that are on some dealer RIM profiles and not on other dealer RIM profiles. These longer ORTs could also result in out-of-stock conditions that could negatively impact fix-it-right-first-time (FIRFT) levels and owner satisfaction. Worse yet, if your RIM program has fairly timid objectives (e.g., covers only a fraction of parts), then a collapsed warehouse network would result in longer ORTs for a very significant population of dealer parts … and this before/after condition would disturb dealer parts managers, screw-up dealer DMS re-order points, force more inventory to dealers, and cause other sorts of enterprise havoc.
Time Out: Gene Metheny commented on this: “I think what your friend is inferring is that dealers will end up stocking different parts, and if you can’t get good parts commonality at the dealer, then it is hard to destock those at the forward warehouse. I agree with her. It may be possible for OEMs with fairly homogeneous products or dealers like Saturn to pull this off; but in general, my clients have been unable to do so. The problem, particularly for CE/Ag businesses is that dealers tend to have a focus on different types of products – Row crop vs Hay and Forage, Big vs Little construction equipment, mining etc … many parts are common across applications but may be fast moving for one dealer and very slow moving for another. With one CE/Ag client of mine there are over 500K active part numbers. If you assume each dealer is carrying from 5-10K part numbers, and their movement classes are very different, it is impossible to build a network strategy around RIM. Another CE OEM has been trying to do this, with a fairly manual RIM system where they negotiate stocking lists with each dealer and utilize emergency order depots to fill the variance. Before RIM, they used emergency order demand to figure out what to stock in emergency depots and found this was extremely volatile. What was ordered on emergency today, could be ordered on stock tomorrow, not because the dealer stocked the part, but only because the situation wasn’t urgent enough to pay the emergency order fees and emergency order freight. They are hoping that this tends to stabilize a bit more with the new RIM program with special EO terms and conditions. We will see what happens. I agree that being able to leverage RIM to modify your network is the golden egg. But I am not convinced it is possible.

Time Out: Harry Hollenberg had this to say: “Gosh, we’ve been looking at this for years and each new piece of data seems to contradict the last. I understand the theory – put dealers on RIM, make sure they have the right parts in stock, consolidate your PDCs and offer more reliable, but longer ORT deliveries for non-RIM parts. As Gene says, don’t think that’s ever been proven to work yet. Here are just some of the findings over the years. Saturn historically had the highest Availability Satisfaction in the Parts Survey, even with one PDC and up to a week ORT. We attributed that to RIM – ‘if dealers don’t actually order, they can’t calculate the response time – they just see a truck showing up on their dock with stuff they need, so they’re happy’. Of course, with only around 50K part numbers, a handful of models, and the closest Saturn dealerships also owned by the same Principal (and typically pooling their inventory) – life was pretty easy for them. Takeaway: RIM reduces need for 24 hr stock order ORT? Parts Survey analysis shows dealers have no clue what their facing fill is, and OEMs with higher facing fill do not have dealers that are more satisfied with Availability. Takeaway: Chasing facing fill may not be the right approach, since it doesn’t seem to translate into happier dealers. Corollary: Whatever your fill rate is, dealers can sense when it goes south (backorders, recalls, SAP) – and then they get really pissed. So increasing facing fill may not help, but decreasing it hurts. BTW, none of 2009 Availability Satisfaction industry BIC OEMs have a RIM system. 2010 CE Availability Satisfaction industry leaders do not have RIM systems. I haven’t done much work in this area lately, but when Mike and I did an analysis about 2 years ago for one OEM, only about 50% of a dealer’s volume was going to be “RIM” managed. That’s a lot of volume (relatively fast movers if stocked at a dealer) that still has to be handled the traditional way. So taken together, I see the benefit of RIM from the OEM’s standpoint in terms of protecting purchase loyalty and maybe even increasing sales (in Harley’s example), but I agree with Gene – I haven’t seen anyone able to get enough volume flowing through their dealers on RIM to justify consolidating the network and slowing down ORT on the other stuff. If the “other stuff” was all non-competitive product, I suppose you could say “tough luck”, ‘cause the OEM is the only option. However, I don’t think it’s all non-competitive stuff. So, it seems that until you can get 1) dealers to stock lots more parts, or 2) dealers to consolidate and share inventory locally (dealer depot??), or 3) significantly reduce your part count by sharing part #’s across models), or, agree to air-freight non-RIM orders (possibly controlled by T&Cs to avoid excessive use by dealers) – it’s going to be very hard to pull back on PDCs.

Time Out: Mike Sachs also commented: “Harry sums it up pretty well – as far as the OEM and dealer are concerned. The one thing that remains unclear to me is the end-customer. The fact is that if the part is not on the dealer’s shelf and he can’t buy it locally same-day, then the customer is either going to wait for the part to arrive or have the vehicle repaired elsewhere. The wait might be one day if the PDC is nearby or the wait might be two or three days if the PDC is farther away (I suspect the minimum wait is really two days, since the dealer is not going to call the customer back until the part is in-hand). Bottom line is that it’s still a wait. Having to wait (for any length of time) will cause some customers to go elsewhere. Having to wait three days instead of one or two will cause some additional customers to go elsewhere (this time - maybe, next time – more likely). Ultimately, this is the customer we need to satisfy.

That’s the best I can do on the “yes” side. Pretty convincing. But, I’m a contrarian on this one (or something resembling a lone Democrat Supreme Court Justice dissenting opinion on gun control; remember, I am in the NRA) Let’s look at the “no” side – the arguments that say facing fill is not that important in a RIM-world … even in timid RIM worlds.
  • The lion’s share of RIM-controlled parts should be competitive parts; RIM should ensure even higher levels of dealer purchase loyalty for these parts. You need to balance higher RIM purchase loyalty with lower non-RIM purchase loyalty. If this is not true with your RIM program, you might want to reconsider your RIM policies.
  • Larger, older, more diverse dealer bodies have many different parts stocking needs – call them RIM stocking profiles. Therefore, RIM suggestions could be wildly different from profile to profile. When you sum everything up, you will find that only fraction of your parts population is RIM stocked … and that small populations are uniformly RIM stocked at all/most dealers. Some portion of the dealers’ receipts will be RIM-controlled and some portion will be traditionally controlled by the dealers’ DMS. That’s life. If your warehouse network has been collapsed to support a uniform RIM strategy – your dealers will experience different ORTs for different parts (sort of like today with increasing ship direct, where a typical dealer order can be delivered in 8-10 different shipments and by several carriers … but, oh my, that’s OK since we save a ton of money on that … never mind). Some daily stock orders will have irrelevant ORTs because they are on RIM; they won’t even look like a stock order. Other “real” daily stock orders will take a few days longer to arrive, consistently, from a more distant facing warehouse. Big deal. OK, if this really is a big deal, I’d scratch further. It may be that the “big deal” of a few days longer on a non-RIM stock order means (1) that your dealers are using free daily stock orders for critical order situations, (2) that your terms and conditions do not encourage proper dealer stocking and use of critical order situations, (3) that your RIM objectives are too timid, (4) that you have not properly trained your dealers to work in a real-world RIM environment, and/or that (5) change stirs up the mud; the mud will settle in time and you will be able to see clearly again.
  • A collapsed “RIM” warehouse network should result in fewer and more robust inventory stockpiles; these should result in higher fill rates inside those longer ORTs, which should offset some of the negatives associated with longer ORTs. If this is not true with your RIM program, then you have very different strategies at work: (1) a RIM strategy, (2) a multi-tier warehouse strategy (where there are “RIM” warehouses and other more normal “facing” warehouses), and/or (3) an inventory-reduction-at-the-cost-of-fill strategy … so don’t blame the lack of fill upside on RIM. Square-root law effect here.
  • The industry is migrating to performance terms and conditions. RIM compliance and in-stock fill ratios should be part of the “performance” in “performance terms” – don’t blame your RIM program’s longer facing fill on not-well-thought-out terms and conditions.
Bottom Line: My inner pragmatist tells me that my brilliant friend who started this all, Gene, Harry, and Mike are correct: RIM does not obviate the importance of facing fill. My other side, the right side of my brain, is exasperated. How did we evolve to a world where taking back a couple of days of ORT causes our retail house of cards to fall? Why can’t we unwind once good decisions that are now bad decisions from the past and substitute better strategies and better decisions? Why don’t we integrate our supply chain strategies with our sales, marketing, and financial strategies? My right side of the brain sees all this and concludes that there’s still a lot to learn from Ford. Oh, :-)

No comments: