Author Archives: thedoctor

Economies of Anti-Scale

Late last month, Mr. Smith reposted a couple of great posts on why Bigger Procurement is Not Always Better Procurement (Part I and Part II) over on SpendMatters UK. While sometimes bigger spend equals bigger discount, this is typically only true for the acquisition of consumables where there is a predictable economy of scale that kicks in at higher volumes — such as the production of identical goods on a production line or the sale of multiple software licenses.

In some markets, as Peter points out, increasing volume decreases costs. Let’s review his examples to understand why.

Short-Term Contingent Labour

Let’s say you want 100 additional workers for 20 days to help you stuff boxes for the Christmas rush. You might think that you should get a much better rate than the 10 workers you hired for your annual 4th of July promotion, but this is not likely to be the case. First of all, lots of organizations in your position will want extra workers for the Christmas rush, and the contingent labour organization will only have so many. Secondly, even if there is demand, why would an organization want to hire resources that will then be on the bench until at least February (for the mini-Valentine’s day rush)?

Bulk Pre-Paid Hotel Room Rates for a Year

Hotels, like airlines, make their money during peak travel and peak conference / festival seasons, and during these times, when they can charge the full amount allowed under law, the last thing they want to do is give you a room for 30% off of the normal rate, which is typically less than 35% of what they could be charging. As a result, the best deal you’re going to get is the average price they got for a room last year, as they will be hedging their bets.

Energy

Most energy plants still rely on oil, coal, and natural gas and, as a result, energy costs are dependent on the somewhat unpredictable prices for these limited resources. Plus, these energy companies can always get the maximum allowable price from consumers and small to midsize businesses with no negotiation power, so if they are selling most of the energy they are producing, and giving a big contract might require them to occasionally buy energy from the spot market during peak (heating/air conditioning) season, your big contract, that requires a big discount, is not attractive to them.

These are all economies of anti-scale. An economy where volume increases uncertainty (such as the short term contract for contingent labour who could be benched for long periods of time if hired by the contingent labour provider or the energy example), decreases profitability (such as the hotel example where having to give you a room during peak seasons cuts profit by 60% or more), or increases overhead cost per unit beyond a baseline is anti-scale. This last case includes any situation where customization is required or limited edition runs are required, as is common in jewelry or toys and games. Customization requires labour, and beyond a certain number of customized units, the provider would have to hire more labour who could not be fully utilized at the time of contract signing. This adds risk, and cost. Similarly, it does not matter how many limited edition collector cowbell* orders of 1,000 you put in if each requires a specialized mould to be produced. Each mould and line set up requires a fixed amount of time and that fixed overhead does not scale with more custom orders.

Thus, when sizing a spend opportunity, it’s important to first identify if you are dealing with a scale or anti-scale economy. There will typically be a much larger savings potential with an economy of scale than with an economy of anti-scale. Thus, if your organization is being measured primarily on savings, it’s important to identify those economies of scale categories as soon as possible.

* Even if you will always need more cowbell!

101 Procurement Damnations – We’re Almost Halfway There

Our last post chronicled our 50th Procurement Damnation that you, as a Procurement professional, have to deal with on a regular, if not daily, basis. If only these were the only 50 damnations clouding your mind and getting in your way. There are still 50 more damnations that are just as pervasive that seemingly exist only to pester you on a daily basis that we have not yet discussed!

However, before we get to the next 50, we thought it would be a good idea to summarize the list to date so that you could go back and review any posts in the series that you might have missed as this is SI’s biggest and most aggressive series to date, longer than both the 15-part “Future” of Procurement series and the 33-part “Future” Trends Expose series (that followed) combined and double the length of the maverick‘s 50 Shades of Pay series (assuming it gets completed) which, to date, only has 10 parts up and available for your reading pleasure.

There’s more that could be said, but much has been said in the 50 posts published to date and much more will be said in the next 50 posts, so, without further ado, here’s the first 50 for your reviewing pleasure.

Introductory Posts

Economic Damnations

Infrastructure Damnations

Environmental Damnations

Geopolitical Damnations

Regulatory Damnations

Societal Damnations

Organizational Damnations

Authoritative Damnations

Provider Damnations

Consumer Damnations

Technological Damnations

Influential Damnations

Bonus Posts!

Technological Damnation 91: Proprietary Madness

It’s bad enough that we have to deal with IP & Patent Madness, as chronicled in our post on the 89th (Technology) Damnation, but proprietary madness is likely to drive us all mad (and may someday push the doctor over the edge, into the land of the crackpot, where at least one blogger in the space is already dwelling).

Just what is proprietary madness? It’s mega-corporations, especially in software and electronics, taking the rights of ownership to extreme. Started, and continued, by the current and former Technology heavyweights, including the likes of IBM, Microsoft, and SAP, it’s not only the creation of company specific standards for software and hardware interfaces, its the restriction of the specification of those interfaces to approved partners and suppliers, limiting the supply of support services and related products to a handful of vendors. This not only drives up the price of those products and services to well above the market average price for support services for software and products with open and published specifications, but can make it difficult, if not impossible, to get support when demand is high or related products if one of the few vendors who can produce products shuts down.

Those of you with SAP know exactly what we’re talking about. Unlike Oracle, which publishes its core schema, and does not change it between minor versions, SAP does not publish its score schema, does not guarantee any stability between bug updates between minor versions, such as between 4.7.1 and 4.7.2, instead requiring you to go through its proprietary NetWeaver interface, which you will, of course, have to acquire to actually support any customizations (and likely build applications in the Portal). And learning the portal is no easy task. One of the most complete books on it is 700 pages alone! Then you need to find the documentation on the data stored in each R3 module you are interested in and how to get it out. There’s a reason that not every shop does SAP support — and that’s because, even though SAP now has a lot of documentation on their website, you need weeks of expensive training just to learn the basics of Portal Development, R3 interface, and the core data types and record types used to pull the data you need out of R3 and push modified data back in. Getting to the point where you are effective at developing and integrating custom supply chain applications requires months of training and mentoring and years of experience. As a result, it’s typically only SAP partners who can provide this support. In contrast, with an open Schema, as found in Oracle and MySQL, all you need is SQL experience and the interface library for whatever language you are using (whereas NetWeaver limits you to Java) — which makes it much easier (and cheaper) to not only find support resources, but vendors with best of breed software modules and platforms that can plug and play with Oracle right out of the box!

But it’s not just software vendors that create proprietary technologies, it’s hardware vendors too. Dell, IBM, HP, etc. all have custom control and administrative solutions for their server platforms. Want a third party virtualization platform to work out of the box on a new server configuration and take full advantage of the capabilities? forget it! You’ll probably have to wait six months to a year or more before third parties, like VMWare, are optimized and configured for those platforms (assuming that the full specifications are published upon technology release and licenses for custom drivers aren’t required), making their administrative software a must if you want to upgrade to the latest technology, and not upgrade to technology that was outdated a year ago.

But it’s not just IT companies that have proprietary technologies and interfaces. Big electronics companies do this too for most of their consumer (and even enterprise) electronics, including companies like Samsung (and its new Mobile AP core) and Sony (and its new ultra high definition TV technology).

And while there is nothing wrong with proprietary technology, as a company needs some assets in order to survive, the lengths at which some companies go to keep it secret and protect it, in a world where data needs to be shared and products need to be utilized with other products makes development (and the supply chains that rely on that development), a nightmare.

We need open standards and open interfaces. The sheer existence of IE alone should make that clear. (There’s a reason that many new IT start ups simply won’t support it anymore, and that’s because they can write stuff that runs almost flawlessly in Chrome, Firefox, and a dozen of other browsers or that runs almost flawlessly in a single version of IE on Windows platforms only, but not both. Since Chrome and Firefox and similar clone browsers run on all major platforms, and IE doesn’t, and since Chrome and Firefox almost fully support the open standards, whereas IE supports the Microsoft standards and those portions of the open standards it feels like, and, to top it off, [older versions of] IE allows case insensitive JavaScript!) Restrictive proprietary standards and interfaces just make life unnecessarily difficult.

But too many companies are too big and powerful, so it’s not going to happen and we’ll be forever wasting countless hours checking interface requirements, versions, and support availability instead of focusing on whether or not the technology meets our needs and will help us get our work done. It’s more daily damnation for all of us.

Today Nintendo Lowers its 2DS Price to Just 99.99 USD

In the ongoing battle to milk every cent out of an obsolescing platform before it’s time comes to an end, today, as per ars technica, Nintendo lowers its 2DS price to 99.99 USD.

Why is this significant? It emphasizes a harsh reality of Sourcing in just about any technology or non-raw material category. Whatever you’re sourcing today, you won’t be sourcing tomorrow, and if you are, chances are the organization won’t be around much longer as sales will dry up, the balance sheet will dip into the red, and bankruptcy will be inevitable.

But should it be this way? In the age of the PC, even though, as Weird Al clearly pointed out in the now classic It’s All About the Pentiums, it was obsolete before you opened the box, it didn’t mean that you had to throw the whole thing out and get a new one to take advantage of advancements. Motherboards had removable processors extra slots and you could throw in or replace cards with math co-processors, better video cards, parallel and serial device interfaces to printers, scanners, and analog signal converters, etc. Now, you knew that eventually you would have to upgrade when a better bus came along or the register size doubled, but even then the new mother boards came with interface slots to the previous generation cards so you could keep using them until you were ready to replace them. You could keep the same case for the better part of decade with smart upgrades.

Now we have slim case laptops where everything is built in and nothing is upgradeable. You have to buy a whole new unit every two years. Not only does this mean Sourcing needs to source a whole new product design at least every six months, but it also has to focus on reclamation. Many modern electronics, especially those that run on cellular or wireless networks, require a significant amount of rare earth minerals and expensive metals that need to be reclaimed due to the limited supply. Plus, in many locales, it’s illegal for a consumer to throw it out, and not only do they need to take the product to a recycling location, but some locales, such as the EU, require the producer to take the product back and appropriately recycle it.

But you know all this, as SI has been ranting about this and the need to design for recycling since the beginning, but, at this point, that’s not enough.

At this point, SI really thinks that all products need to be designed for perpetual upgrade. It should be possible to replace all components of a device as needed as they wear out or need to be upgraded. And it needs to be easier than it was with old desktop computers where you had to open the case, remove a bunch of wires to get to the card/drive/processor, do a precise sequence of presses, twists, and pops to safely get the component out, do the reverse to get it back in, reattach the wires, put the case back on, and then power up and test you don’t cross any wires (while wearing rubber gloves, just in case).

Each component should be a self contained “box” with a standard interface connector, using an upgradeable design that can support the fastest speed the configuration of connected components can effectively support. For example, a copper-based multi-pin connector (which, as demonstrated by Thunderbolt and USB 3.1, can support data transfer rates in excess of 10 Gbit/s) for low-end consumer devices and optimal connections (which, as multiplexing technology, will allow faster and faster transfers in the future) for high-end consumer devices and business devices. Boxes should have multiple smart connectors that can register the type of device they are connected to, and the devices they are connected to (as the components will communicate over an internal high-speed network), allowing the device to be upgraded with new components, and capabilities, not imagined when the initial set of components were first built.

For portability, durability, and weather-proofing, custom enclosure boxes could be built that would hold a standard set of components that would represent a power-house desktop computer or a portable tablet/laptop (where the screen slid out of a sheath and plugged in to the top of the main box and the keyboard folded down).

We may never see this, but imagine how much easier it would be for everyone if the same components could be used for years, investments lasted longer, and Sourcing strategies could be more consistent and predictable.

Just a revelation encased in a rant triggered by a reaction to another price reduction required by planned obsolescence.