Friday, July 24, 2015

Lets Talk Facing Fill – Carlisle Scrapbook
by David P. Carlisle

Parts people can be very focused on “parts availability” as a key metric. The big issue is: what does availability mean to a smart dealer parts manager?


Every year, we survey thousands of dealer parts managers, and occasionally we’ll follow up the survey with individual interviews. Recently, we conducted a few such interviews with parts managers from leading OEMs to ask them this question on availability, and what they told us matches insights from our previous data analyses and what we have heard from some OEMs. What follows is based on one interview that is representative of the parts manager interview group.


OK, this seems simple enough. But really, it’s not. Here’s what we found:
  • When we asked what does “availability” mean? The response was: “What I can get from my warehouse in a timely manner. Or, whether I can get the part at all.”
  • The parts manager reported that their facing fill was at 94-94.5%, and said he’s very happy with this; said they haven’t been in this good of a position in years. He says network fill is something like 98%.
  • He is very satisfied with his facing fill rate and thinks there are more important issues that the OEM could focus on to improve his business.
  • Unless a customer specifically asks for something else, he will always order a genuine part. Essentially, only if the customer says they cannot afford the genuine part will he go aftermarket – he will try to convince the customer that genuine is better than aftermarket (better quality, fit, performance, and warranty).
  • How loyal was this parts manager? He indicated that substantially more than 90% of his parts come from the OEM.
Like a Talmudic Scholar, I’ve put a lot of thought into the 164 words in these notes. What do they really mean? You can see how I marked it up, parsed the words, linked the clues, and squeezed out the essence of what the parts manager said.


At its essence, “availability” is about the level of backorders: “whether I can get the part at all.”


Toyota got it right decades ago when they adopted a goal of 95% facing fill. Back then, I asked Toyota Motor Sales all-stars Bill Bucher and Ace Yeam where the 95% came from. They had been Korean War supply sergeants in the Marines and said that’s what they aimed for in the military. But, a lot has changed since the 1980’s. Facing fill was simple back then – it was the percent of “perfect” lines filled on all dealer orders. By “perfect,” I mean completely filled order lines – if a dealer ordered 100 washers and the parts warehouse could only fill 99, this order line was deemed “unfilled.”


Now, facing fill is not so simple. The order might be sourced from multiple warehouses and supply points – bulky fenders might come from a high-cube center, while expensive, small-tech parts might come from a central, low-cube facility; slow moving parts might come from a slow moving warehouse or Vintage, and other parts might come from ship-direct suppliers.


So, what is facing fill? It is the percent of order lines that arrive at the dealer inside the expected order delivery time. It should be about 95%.


Because of the incredible evolution in the parts supply chain, “facing fill” might not be all that important any more, other than as statistical evidence of appropriate parts availability. It’s like runs, balls, and strikes in baseball. There are a lot of different ways to make a run, to hit a ball, and to throw a strike. But, who really cares? When you turn over the picture of a baseball card, it’s all about simple numbers: runs, balls, and strikes. There is also a lot of ways to achieve what a dealer interprets as 95% facing fill, and all that really matters is that 95% of the parts are there when they expect them to be there.


Bottom Line: So, what really is availability? To a dealer, it’s not so much about what’s available as it is about what isn’t available. The best representation of this, it seems, is backorders.


Friday, July 17, 2015

Why Old People Are Good For Us – Carlisle Scrapbook
by David P. Carlisle

The 2016 Customer Sentiment Survey** looked at Net Promoter Scores (NPS)1 across a bunch of different criteria. We found out something that we pretty much knew already – older dealer service customers are more bullish than younger dealer service customers on their brands and their dealers. Millennials (18-34 years old in the chart) have a NPS brand score of 36.9% and NPS dealer score of 24.4%. Boomer NPS scores are nearly twice as high. See, old people are good for us.


We really need those old people, because when we look at Net Promoter Scores by type of service, we see a pretty sharp falloff when service customers only go to dealers for “repairs.”


I think it is logical to conclude that since boomers, who are older than millennials, are stronger “promoters” of their brands and their dealers, they will hate their dealer less after they go for repairs.


This is important, because dealers have been losing out to the aftermarket for years on non-warranty customer-pay maintenance and scheduled service work. What’s left in the “tube of toothpaste” is tough-to-do, expensive, repair work.
Timeout: Given that 1) boomers tend to have a more favorable opinion of dealers and brands, and 2) maintenance visits result in higher dealer/brand NPS scores than repair visits, one might be inclined to think that boomers simply have a higher tendency than millennials to use the dealer for maintenance. However, our data suggests that there is no difference between age groups regarding the proportion of dealer visits for maintenance vs. repair work.
Bottom line: We are losing our millennials. Dealer service can be corrosive to a customer relationship, especially with young customers. All of our plans won’t change this … maybe some of the reason for plan failure is that they are typically developed by older boomers who have brilliant insights concerning what they can really accomplish inside their company and with their dealers. That’s why everybody loves Elon Musk and Tesla. He takes giant steps.


NPS’s for service customers covered by a pre-paid maintenance plan show that they are stronger “promoters” of their dealership and their brand. This makes sense, because the pre-paid maintenance plan foots the bill for the work done and nicely sidesteps the service customer’s top problem with dealer service … that it costs too much. Here’s a baby step for you. OK, it might not be a baby step, but pre-paid maintenance is pretty much all upside. Push it. This is certainly less corrosive than the status quo.


Might be worth a try. Well, do you feel lucky?


** The Customer Sentiment Survey (CSS) has been around since 2007; its objective is to better understand service customer needs and opinions. From 2007-2014, the survey was predominately cross-industry in its focus. In 2015 the survey started deep drills on brands. The CSS is not a beauty contest survey; its primary purpose is to better understand service customers so that OEMs can better give them what they want and need. Further information is available by contacting Thomas Neumann at Carlisle & Company tneuman@carlisle-co.com.


_______________________________
1Net Promoter Score is a customer loyalty metric developed by (and a registered trademark of) Fred Reichheld, Bain & Company, and Satmetrix

Thursday, July 9, 2015

What Good Are Net Promoter Scores? Carlisle Scrapbook
by David P. Carlisle

Various online publication sources provide NPS scores for top-of-mind brands. Clustered close together are Acer, Audi, BMW, Amazon, and Apple.


What do all these brands have in common? A Net Promoter Score (NPS)1 , which is a measurement of the loyalty that exists between a provider and a customer. It is based on a simple question:


How likely is it that you would recommend our company/ product/ service to a friend or colleague?


The scoring is based on a 0-10 scale, where the 0-6 respondents are considered “detractors”, 7-8 respondents are considered “neutrals”, and 9-10 respondents are “promoters”. The Net Promoter Score is the percentage of promoters minus the percentage of detractors.


So, you are at a barbecue and you ask, “Hey, what do you think of Amazon?” Your friend replies, “I think you should try it …” Your friend is an Amazon promoter. Next you ask, “What do you think of your BMW?” He responds, “I love it, you should get one.” Now, he’s a BMW promoter. The answer to the question assumes some sort of familiarity with the product.


OK, now you ask, “What do you think of McDonald’s?” The friend replies, “Not a fan, I do not recommend it … you should try Five Guys.” He is a McDonald’s “detractor.” We have jumped from the web market, to cars, to quarter pounders. Can there possibly be a linkage between these three?


I think so.


Specifically, the linkage is all about word of mouth reputation. The friend recommends (stands behind) Amazon and BMW, but not McDonald’s.


I, for one, am convinced that this linkage is relevant.


The next question is how relevant are the NPS scores? The gray chart shows some NPS scores that are easy to get from the internet – npsbenchmarks.com. Here we see the lineup of some familiar automakers … and Amazon, McDonald’s, and Facebook.


The naive read of this chart is:
  1. Tesla has more promoters than Amazon; an inference that is easy for me to accept.
  2. Harley and Honda have more promoters than mainstream automotive brands. Again, easy to accept.
  3. McDonald’s has fewer promoters than Amazon and the auto companies. Makes sense to me.
  4. Facebook has even fewer promoters than McDonald’s. Interesting.
So, I can accept the simple inference of the numbers and how various brands line up. But, what do they mean in an absolute sense? We need to go to the ultimate spaghetti western for this – The Good, the Bad, and the Ugly. Amazon is a “good” benchmark for NPS – it sits at around 64%. McDonald’s is at negative 8% (-8%) and that’s pretty “bad.” Facebook is the poster child for “ugly” at a NPS of negative 21% (-21%).


Using these benchmarks, we see that the auto brand NPS scores are neither “bad” nor “ugly.” Facebook and McDonald’s are doing pretty well and neither seems to be on the brink of disaster. So, what do bad and ugly NPS scores mean?
Time Out: For established brands, I think of a bad/ugly NPS as a form of corrosion. Something’s bothering customers. If left unremedied, they will ultimately flee. This is easy to imagine with both McDonald’s and Facebook.
Bottom line: Net Promoter Scores are here to stay – they make sense … and it makes abundant sense to compare one brand with another and have a common, transcendent, measure of word of mouth. Yes, we need NPS benchmarks so that we can take the pulse of brands, products, and services and determine if they are healthy or in cardiac arrest. We need to be able to look at NPS scores and get some idea of “corrosion” that may be present. We need to look at different slices of the market that have different levels of corrosion so that we know where the problems are.


Let me leave you with a vivid example. Budweiser’s NPS taken from the same web source in the bar chart is 29%. Not great, but not bad or ugly yet. I encourage you to visit a large store that sells beer in the United States. Go to any one of them. You will find the aisles jammed with hundreds of different craft beer products – squeezing Bud’s shelf space. I suspect that if you could find a NPS survey that asked respondents if they would recommend craft beer, you’d find a NPS score much higher than 29%. The score, possibly, could be as high as Tesla’s.


From burgers, to beers, to Tesla. Hmmm. They might not be all that different. In the next blog we’ll take a look at NPS scores for dealers and OEMs in relation to the customer maintenance/repair experience.

_____________________________________________

1Net Promoter Score is a customer loyalty metric developed by (and a registered trademark of) Fred Reichheld, Bain & Company, and Satmetrix

Tuesday, June 23, 2015

Why Ford Is Brilliant – Carlisle Scrapbook
by David P. Carlisle

Thirteen years ago, in 2002, Ford announced Daily Parts Advantage (DPA) to the world. Back then, it seemed simply brilliant. Now, more than a decade later, it seems even more brilliant.
DPA was an integrated strategy of revised stock order cycle times (“daily”), revised terms and conditions, and revised network structure. Perhaps the cleverest aspect of DPA was looking at parts very differently than anyone ever had previously. Ford concluded that parts were of three different types : (1) parts that needed to be close to dealers for a 24-hour cycle time and 24-hour order response time (ORT), (2) small parts that could be easily shipped by a parcel carrier and beat a 24-hour ORT, and (3) large bulky parts, typically used in collision repairs, that required a sparser network and could easily flourish with a 48-hour ORT.


Seems obvious. But, it sparked a lot of debate … particularly concerning the large bulky parts. Did dealers need them in 24 hours, or was 48 hours sufficient? Would collision market share suffer with longer ORTs? Would insurance companies react to longer ORTs that could ripple over to longer rental contracts for the insured? After thirteen years, the answers are in:
  • Nah
  • Nah
  • Nope
Collision market research that we are currently conducting confirms all this, and more. Ford was right and kudos to their team of innovators: Don Johnson, Kent O’Hara, Frederiek Toney, Anu Goel, Joe Kory, John Sullivan, Helmut Nittman, Brad Wallis, and many many more. Cisco Codena headed up aftersales at the time and his leadership (call it courage) deserves mentioning.


The collision research we are doing spans a wide area – the primary objective is to refine our estimate of the impact driverless vehicles and collision avoidance technologies will have on collision parts sales. However, one work stream involves process mapping typical current-state collision repair facilities. So far, we’ve seen three things that are interesting:
  1. Price. Because the insurance industry largely controls what goes on in a collision repair shop, the advantage always goes to the low price option. This explains why non-genuine parts dominate repair orders for “unprotected” vehicles. In some states, laws protect some owners from repairs using non-genuine parts. However, these laws typically only cover only newer cars and not all specify “genuine.” States with significant “genuine parts” protective legislation cover only about 22% of the population. Outside that, it’s the wild, wild west.
  2. Price. Order lead-time differences between genuine and aftermarket (AM) parts are not much of a deciding factor when sourcing a collision part. For “competitive” parts, OE order lead-times are generally competitive with aftermarket parts. However, aftermarket return rates are ~30%; almost six times the return rates of OEs. Factoring in this return rate differential (in order to receive a correct part) the expected cycle time difference between an AM and OE nearly vanishes. Problems arise with “non-competitive,” slower moving, genuine parts that have erratic lead-times. Since they are “non-competitive,” this is not much of an issue . I suspect that there’s never been a successful aftermarket company that decided to invest in tooling simply because they thought that having inventory and quicker-than-OE lead-times would make a difference. No, the decision is always made on volume and price.
  3. Price. OEs’ aggressive collision parts “price matching” programs might not cut it. In a case study of one collision shop, the owner took advantage of every price matching program offered by OEM dealers. Most of the OEM parts were bought at either an extended 11% discount off the already reduced OE part list price or a 30% discount off the independent aftermarket part list price. At the end of the eight months, the shop analyzed the results: labor profits increased 3 points, while whole parts profits decreased 12 points due to the price matching programs. The part profit impact is a pure, measurable loss, while the labor gain isn’t even necessarily absorbed by the shop. Overall, shop profits dropped by 5%. So, while collision shops readily admit OE fit makes for timelier repairs, the benefits do not drop to the bottom line.
So, yes, Ford was brilliant. Carving out separate supply chain strategies for different classes of parts saved money … and worked well in an integrated
strategy that led to huge improvements in dealer and end-customer service.
(Most of these customer benefits were encapsulated in Ford’s move to a daily stock order.) Our research, so far, seems rock-solid in supporting what Ford did with their bulk parts centers – because the big issues in the collision market are price, price, price. Marginal advantages in lead-times have about the same lasting market impact that larger tail fins did on late 50’s sedans.


Bottom Line: That was the good news. The collision market is steadily becoming even more price driven, and Ford got ahead of the curve here. The bad news is that the insurance companies have been busy and have effectively commoditized the market during the past few decades. Price is more important than ever, and the squeeze will continue to tighten. Most of the genuine collision parts are used in the newer car parc, which is still continuing to recover from the 2008/9 recession. This makes for pretty nice year-over-year sales gains in this segment. However, when we finally achieve a steady state that resembles pre-recession conditions, we are bound to see that the OE segment of the collision market has shrunk. Big time. Then, when we lay on the anticipated market-melting impacts from collision avoidance technology and driverless vehicles … as my grandmother in-law would say, “oy veh!”

Tuesday, June 9, 2015

“Small Data” and Dynamic Scripting – Carlisle Research Scrapbook

“Big Data” is a big thing in aftersales service support – it will make our lives better in inventory management, sales forecasting, parts and service pricing, … and, ultimately, in better serving the needs of service customers. But, it’s going to take awhile.


Sorting through blizzards of data, big data is a valuable predictive tool. But, when the sightlines are fairly clear, we don’t have to put a lot of time and effort into fancy models and monster analyses. Effective service processes and simple surveys can capture the data directly from the customer - sample size of one - and be 100% accurate.


We might start with some recently available “small data.” We are about to release the syndicated 2015 Consumer Sentiment Survey – Dealer Customer Wave. Let’s take a peek.


#1. Some people will wait a long time, others won’t wait at all, yet others (more than a third) will wait an hour. Knowing the customer’s urgency is important information when prioritizing service, offering substitute transportation, and dispatching jobs to the service bays.


#2. Some people want to be greeted at their vehicle; others prefer to be greeted at a more traditional service advisor’s desk. Knowing this is important in designing dealer service lane processes – but again, “small data” shows us that even within a single dealership; there is no one size fits all.


#3. Not every service customer has the same preferences regarding how they are contacted for status updates. Less than 40% prefer “modern” methods – about the same as those who prefer phone calls. Again, different people are different.


These simple insights come from “Small Data” – a few thousand surveys of individual brand-service customers. The insights here show us that there isn’t one right answer that fits all – the data steers us into looking at customers within discrete segments, each segment having discrete preferences. We all know this is true.


So, why do we treat each service customer the same way when they schedule/arrive for a service appointment? It makes no sense.


Bottom Line: The technologists who are designing the tablets for Service Advisors should rely on some “small data” to accommodate service customer segment preferences. The service experience should be dynamically scripted to maximize accommodation of consumer preferences, which should then maximize consumer satisfaction.


To be perfectly clear, if I know Mary-Joe’s preferences are that: (1) she will wait up to 1.5 hours, (2) likes to be greeted at her car, and (3) wants texts to keep her up to date with her service, then I can…
  1. Prioritize my workload to make sure her car is finished inside 1.5 hours and not have to give her a loaner,
  2. Give her a priority customer card and shift her to a desk-less service lane, and
  3. Set up a texting schedule that keeps her apprised of progress.
We can accomplish results more readily than with big data by doing simple things like capturing customer profiles at first appointment, at service handoff, at scheduling, or at the owner portal. These can be done up front and occasionally refreshed. This removes the margin of error of big data, because we know we are right. And it’s a lot cheaper.


And, most importantly, we can deliver to Mary-Joe a flawless, repeatable service experience.


Wednesday, June 3, 2015

The Connected Vehicle First Needs To Be Connected – Carlisle Research “Scrapbook"

We do a lot of research about what goes on in the dealer service lane. Sometimes it is interesting to connect the dots between disparate sources of data and see if it better explains life, or what life promises. Here are three pictures from our scrapbook and some simple messages. What do you see in the scrapbook?
  1. According to dealers, not many repair orders are scheduled online.
  2. According to customers, many prefer to schedule online (more than those actually doing so, as shown above), while many others prefer to just call in and schedule their service.
  3. It seems that most customers would prefer to connect from their vehicle to schedule service.
Bottom Line: It seems that we have a problem with the tools we provide customers to schedule service today. Many are interested in doing it electronically, but less are actually doing so. Either dealers are slow to adopt or the current set of web-based service schedulers are flawed and not liked much by customers or dealers. Beyond this, customers want the convenience of scheduling from their vehicle, but we still have some work to do here.

Thursday, May 21, 2015

Why Driverless Vehicles Are Inevitable, and Why We Care About This
by David P. Carlisle

For quite a while, driverless vehicles have been in the news almost daily. Now we hear that Google has perfected its self-driving car, and that driverless vehicles will be commonplace inside the next five years. This is very important to everybody in our industry, as well as to consumers. Safety and cost are the two big benefits to vehicle owners and drivers. Driverless vehicles will crash less, which means fewer injuries and fatalities, and, inevitably, lower insurance premiums.


For the industry, fewer totaled vehicles will take a bite out of vehicle replacement sales. Lower collision rates will take a bite out of collision parts and labor sales. The driverless car will profoundly change the industry. So, yes, we care about it.


In fact, this innovative “product” is almost fully baked and ready to go. As a successful prototype, we already “have” it. But, do we “want” it??? Is our nation’s legal and regulatory apparatus ready and willing to accommodate this new technology ?


I think those changes will be made. Here are six good reasons why the driverless vehicle is inevitable.
  1. It’s not that big of a deal for our highways to accommodate driverless vehicles. In the U.S. we have approximately 4 million miles of roads. About 3% of those lane miles are devoted to interstates/freeways/expressways, but that 3% also carries about 30% of all vehicle miles traveled. We predict driverless vehicles will probably be using these high-capacity roadways.
  2. We are running out of highway capacity. Our current population of 325 million has grown 14% since 2000, and will grow another 4% by 2020. But, our highways certainly haven’t grown 14% since 2000. In other words an increasing population is being serviced by a highly constricted highway system now operating over capacity. That’s a big problem.
  3. Furthermore, a highway’s carrying capacity is not a smooth linear progression. We’ve all experienced this. You are traveling steadily down a busy interstate, and suddenly see a police car on the side of the road flashing like a Christmas tree. Everybody slows down, rubber-necks, and what was OK traffic becomes a parking lot. This chain reaction – slow human response time coupled with the human capacity for distraction – screws things up. Let the machine drive and the humans gawk, as is their nature. The driverless car would have a very significant, positive, impact on highway capacity.
  4. Boomers are aging. Right now Baby Boomers, the 50 and 60-somethings, represent a tad more than 27% of the U.S. population. By 2020 Boomers will still account for around 20% of the population, but they’ll be slower and their reaction times more retarded. The adverse impact on highway capacity and safety won’t be trivial. Instead of a bunch of young NASCAR drivers, we will have 14% more older drivers with typical declines in vision, judgment, and response time . Yuk. Hmm… if a Toyota can parallel park itself better than a 20-something can, who knows how much a machine can improve on the driving of a 70-year-old Boomer?
  5. Driverless is more green. The energy crisis has melded into an environmental global warming crisis, and it is environmentally irresponsible not to try to reduce fossil fuel consumption. Driverless vehicles, especially in trucking, can take a huge chunk out of fuel consumption through platooning and by taking advantage of vehicular drafting. This works for cars, too. Here, again, machine beats man.
  6. Truck “Platooning” saves lots of money. Imagine a convoy of seven trucks. Typical cost of operation is a little more than $1 a mile for a long-haul vehicle – so this convoy costs about $7 a mile to operate. Around 20% of that cost is for fuel, but fuel consumption can benefit from drafting – maybe we can save 20% of the fuel cost with this. ($1.40 per mile in fuel cost for the seven trucks goes down by $0.28 – a 4% drop in total operating cost.) Labor is about 30% of the operating cost. Typically, labor would account for a bit more than $2.10 of the $7 a mile. But, if we “platoon”, we only need one driver in the lead. So, we save $1.80 of the $2.10 truck driver expense, or another 26% of the total cost. This is serious, compelling, money.
Bottom line: It is a very safe bet that we will see driverless vehicles somewhere on the road inside the next five years.