Wednesday, March 18, 2009

Everything You Wanted to Know About ”Revenue Management” Technology ..But Were Afraid to Ask - Sanjay Seth

Last week we talked about the need to unlearn what we think we “know” about pricing, and to focus efforts on simple creativity to start ringing the bell with revenue management. Let’s focus on these no-brainer simple solutions to revenue management.

Can You Handle the Truth? Or, Are You Lost In Detail and In Denial, With No Enabling Processes and Tools.

When it comes to revenue management, most companies face several hurdles. To start with, most are lost in detail. A typical OEM has thousands of part numbers, millions of transactional dealer data and deep folklores about pricing. Add to that volumes of ad-hoc market research, which typically collects dust. Let’s not forget about supersessions, which at most organizations is a pricing black hole. Beyond that there are sacred cows - part numbers considered “untouchable” without any rhyme or reason. These are all symptoms of either nonexistent or non-explicit pricing strategies and enabling processes.

Organizations also fall into the trap of instituting data-intensive approaches to service revenue management (such as elasticity, transactional analysis, etc.), only to realize that these approaches work in other industries but present problems in service-parts.

Then there is the belief that all pricing problems will be solved once new software is implemented.

The list goes on. You get the point.

Let’s face it. Parts pricing is often called upon to fill revenue or profit gaps, reflecting a visceral sense that there is money on the table that can be added to the bottom line. However, most pricing processes are not designed to handle these issues. There is little visibility as to where the opportunities may lie among an amorphous population of 200,000+ parts, and there are no tools to quickly execute more targeted strategies for realizing those opportunities.

Good Segmentation Is the Foundation

Segmentation is the backbone of service-parts revenue management. Segmentation is all about organization, in that it groups parts into logical pricing buckets. Think of it like Stephen Covey would – “The Seven Habits of Highly Effective Pricing.” Unless you are the “Rainman,” it is daunting to figure out how to set parts prices for 100,000+ part numbers. This is beyond comprehension (hence, the “Voodoo” we talked about last week). However, we can comprehend, organize, and execute revenue management for various groupings of parts; for example, the fast-moving high-wear parts. To start your approach to effective segmentation, you must begin by unlearning everything having to do with how you segment today. If you don’t do this, you will start nodding your head and believing, “Hmm, well, we do it that way, too … just a little differently.”

Segmentation also allows OEMs to overcome the inherent bias towards under pricing competitive parts. Let me elaborate…Cognitively, we think of competitive parts as a single homogenous part group and tend to price these parts similarly and aggressively using the lowest common pricing premium, if any. However, in realty even within competitive parts, the degree of competition varies depending upon the product lines (e.g. hypercompetitive, competitive mechanical parts, tires, glass, etc.). As such, we should price discriminate even within competitive parts, harvest wherever possible, and not leave money on the table.

A good segmentation plan enables you to focus your limited human capital on the parts that matter. Let me explain…For a typical OEM, 80% of the revenue comes from 20% of the parts (fast movers). You should spend 90% of your manual resources on these high-volume, competitive parts. Conversely, you should let technology manage the slower moving, less competitive parts that make up the remaining 20% of revenue.

The key to effective segmentation is defining a set of simple business rules that are derived from existing data. Here are some do-it-yourself tips:
  • Start with three broad categories – Competitive or Market-Driven, Internal Objective-Driven, and Less Competitive. Further divide these based on your specific needs. For instance, hyper-competitive, competitive-mechanical, collision, tires, etc.
  • Identify quantitative filters to drive segmentation. Consider existing part hierarchy, part age/lifecycle, part velocity, level of competition, among others. Add wholesale/retail focus if you want to be fancy.
  • Put high warranty parts in their own segment. This is a must do internal segment. Others to consider are- recall/campaign parts and parts on specific contracts.
  • For competitive segments, use empirical information and heuristics. Focus on fast-moving parts in product lines, since the aftermarket goes after the fast movers as well, rather than the tail end of the sales.
  • Since a part can only belong to one segment, establish priorities for assigning segments. For instance, the high-safety segment will take precedence over the less competitive segment, meaning that even if a part is less competitive, if it is a high-safety part it will be belong to the safety segment.
  • Establish lifecycle triggers that pricing algorithms can execute (and people can review for exceptions) to migrate parts from segment to segment. One such trigger could be volume. Additional triggers to consider are competitive matching, market share changes, etc. This is a must (though it may be not in the beginning if you are just getting started). Otherwise you will be stuck with static, Stone-Age segments.
By the way, all the data you will need to get started is your part master with detailed part information, 12 months of part sales history, and off-the-shelf tools – but more on this later.

Market Data Is Your Achilles Heel, But There Is Light at the End of This Tunnel

In my mind, we make gathering competitive data more complex then it needs to be. Start with your top volume movers and work your way down – you will be surprised how few there really are. Then identify an internal resource (maybe an intern) to coordinate research on these parts (that person should have some people-smarts to get to the competitive intelligence you need).
  • For mechanical parts, send the list to your regional offices and the field force (if you have a really smart graduate, he/she will build a web page that the field force can logon to and punch in what they find). Walk the streets, work the phone, and watch the web. Then, repeat monthly or as directed. Focus on two specific price points; the jobber net and retail price. For jobber net, don’t forget the undocumented discounts, which typically range from 10 to 20%. Also, don’t be overwhelmed with multiple color street sheets (blue, green, etc.). Nobody uses them anymore.
  • If you like technology, invest in an online web scraper. They are cheap and get you published retail prices for a large basket of parts quickly. These are typically good for maintenance-type parts. But beware of retail price, they generally don’t mean anything, unless you are in Europe where retail prices are the key.
  • You can even add a quick and dirty web page or portal for field staff and dealers to input competitive prices with some rewards for those who send good data.
  • If you want to get really fancy, hire a price research bureau. This tends to lend credibility, but generates a lot of data that you don’t have time to digest or use.
  • For collision parts, buy estimating system data – period. There is nothing better out there.
Each Segment Requires a Specific Pricing Strategy and Pricing Method For each segment, define a specific positioning and an explicit formula. I also recommend documenting these strategies using a standard template.

High Warranty Parts – Be careful here and handle them separately. Don’t lower prices, otherwise you may get sued. Self-evidently, treat recall parts like warranty parts.

Competitive Parts – Price competitively (i.e. within the ballpark of the competition), but at a premium. Competitive collision pricing is somewhat different and more complex – you are the market leader, insurance companies are the customer, and there are a few very strong competitors (LKQ, LKQ, and LKQ), all of which make this a little more tricky. You do want to invest some time to formulate a strategy that protects this cash cow.

Less-Competitive Parts – There is not really a market in the classical sense for these parts, so there is really no “good” price. For these parts, use a simple margin and cost relationship: cheap part (in terms of acquisition cost) – higher margin, expensive part – lower margin (you set the cost brackets and assign the margins). You can add sophistication by having separate curves for specific part categories. Scared? Well, if you’d like to make sure you are not completely off base, then look at your key comparable OEMs – but a sufficiently large dataset of comparable parts may be difficult and very expensive o come by. And ultimately, who cares? This stuff is never going to make it into the cost of ownership comparisons anyway. If it does, then your segmentation is not right, as these parts should belong to a cost of ownership segment. Hey, it’s a recession – why waste your money on data that is irrelevant and proves nothing?

Collision Parts Need Different Treatment

More than any portfolio, collision cannot be managed by “averages.” Aftermarket competitors cherry pick the parts on which they will compete. Therefore, we need to manage the collision business at a part-number level.

This requires part-number-level market intelligence from collision estimate data provided by companies such CCC or Mitchell.

Using the detailed part-number information OEMs can identify where opportunities exist for improved margin and/or market share, not to mention better segmentation.

Proper Tools Are a Must In Order to React Quickly and Flexibly to “Contribution Requests”

Start with offline tools before investing in heavy-duty IT. To that end, hire someone who loves Excel and Access and knows some programming (maybe a smart college intern). Have him/her assemble the key inputs, using whatever you can easily get:
  • Parts Master – description, product line info, part creation date, and whatever else is readily available
  • Part Supersessions – old part, new part, supersession code
  • Sales – split by warranty/non-warranty, $ and pieces
  • Cost history – whatever is available
  • Dealer Net history – whatever is available
  • Retail price – whatever is available
  • Application data – which models, model years does a part fit on
    • One OEM built a CD (a simple CD!) that they called the “Universe” with all that stuff on it, published every month by someone in IT using Excel and Access.
Then, have him/her build a scrappy Access database that downloads and integrates all these inputs. The database will maintain segmentation rules, identify supersession chains, run Paretos, and spit out segment codes for each part. Add some reporting capabilities, such as volume and price, either on a part number or product-line level – nothing fancy. You can also store pricing methods and formulae in the Excel or Access database to generate price lists. Also, we don’t need fancy elasticity-type analytics, because there is very little elasticity in service-parts. If it isn’t broke, don’t fix it! The key outputs you are after include volume changes, margin compression due to cost changes, new parts coming on file, etc. Whatever millions you have saved from not building a big IT system, invest into competitive research. Take my word…the Access database will get you going in weeks if not days.

Let me say it one more time…you don’t need technology to get going. If you believe in the crawl, walk, run approach, then you need technology only to run after you have figured out how to crawl and walk using simple offline tools.

In case you want to really invest in technology, I recommend the following functionality in the software system, keeping in mind that very few vendors have demonstrated the experience to deliver this:
  • Data Interface and Management: Integrate and manage various internal and external data sets, including creation of pseudo data from the existing information.
  • Market Research Data: Ability to import and process market research data by part number, including display and use of competitive analytics and various pricing activities.
  • Supersessions: Ability to identify supersession chains from the part data and make it available for various pricing functions.
  • Segmentation: Ability to create and manage revenue management segments over the lifecycle of the part. Segments are the core of market intelligent pricing and revenue management.
  • Pricing Methods: Ability to create, manage, and apply mathematical or logical expressions, formulae, parameters and constraints for determining the precise price of the part.
  • Analytics, Simulation, and Price Generation: Functionality to simulate pricing scenarios, generate price lists, and conduct ad-hoc pricing analysis.
  • Margin Visibility: Functionality to show various margin components at part number level.
  • User Roles and Security: Functionality needed to define, configure, and execute user access to various features of the system based on roles & responsibilities.
  • Alerts and Workflow: Ability to create and manage pricing alerts and workflows, including automatic and batch events and approvals.
  • Reporting and Monitoring: Onscreen and offline pricing/revenue reports (standardized and customized) with hyperlinks to detailed screens/dashboard. Also includes ability to create/deploy standardized and customized reporting and monitoring dashboards.
  • Global Application and Scalability: Functionality needed to deploy and use the system globally, across multiple markets.
Let the truth be told. Today, most OEM processes are focused on reactive price administration. What we need is to shift the focus of our precious resources toward proactive price determination using segment specific pricing strategies, enabled by simple-to-use technology. This will make us strong and profitable as we come out of this recession. Call me at 248-767-2620 if you have any questions.

No comments: