Category Archives: rants

Probabilistic Chips ARE NOT Going to Improve Your Supply Chain Software

It simultaneously humours and scares me every time non-technical folk decide to write about a new piece of technology and how it is going to revolutionize whatever domain they regularly write about. The latest example is this recent piece over on Supply Chain Brain that says one should look for breakthrough technologies this year.

The article, which correctly notes that adoption of ERP systems led to:

  • Stumbling
    as a lack of depth in planning functionality in ERP systems did not lead to integrated planning
  • Failed Promises
    as the ERP is not the de-facto data model for the enterprise or even the end-to-end supply chain
  • A Lack of Agility
    as it has failed to deliver any sensing capabilities that would drive supply chain agility

went on to say that 2011 would be the year when breakthrough technologies using probabilistic chip logic, parallel processing for near-real time response, and artificial intelligence would hit the market. WHAT THE FRACK? Are you kidding me? Did Supply Chain Brain really publish this? Was it written by the Scarecrow from the Wizard of Oz? Let’s examine these technologies in more detail.

  • Parallel Processing
    Your average solution is already taking advantage of this. It’s called a multi-core chip which has been standard in every server for years now. Sure most applications are not written to be multi-threaded or take advantage of multi-cores, but, in order to allow the developers to handle increasing complexity and code-sprawl, most applications are written as multiple modules that are assigned their own processes, and the OS will balance the processes among the cores to speed up overall performance.
  • Artificial Intelligence
    We have been promised (true) AI for 55 years and it has never materialized. What makes you think the next 55 are going to be any different? And how are supply management applications going to deliver a technology that does not even exist yet?
  • Probabilistic Chip Logic
         Obviously the author has no fracking clue what PCL is. Because if the author did, the author would know that PCL, by its very definition, CAN NOT improve computational results. In fact, what PCL actually does is WORSEN computational results. (It will make a decision optimization model worthless since optimization requires millions of calculations and the propagated error would soon be so bad that there will be no accuracy left.) In some applications, like video and audio compression and decompression, small losses in accuracy are not only acceptable but often unnoticeable. It turns out that if you are willing to accept small losses in accuracy, you can do the computations with significantly less energy.
         Most of the voltage required by a computer chip is used to overcome the electrical “noise” created by constantly moving electrons in the chip materials. If this noise is not drowned out by a high enough voltage, then a chip may not be able to accurately determine if an electron flowed through one of its transistors. (Chips produce their bits, 0s and 1s, by measuring the absence or presence of an electron in a transistor.) If the voltage is decreased, the signal-to-noise ratio decreases and the probability of registering an incorrect bit increases. It turns out that the nature of electricity means that voltage (and energy) requirements can be significantly decreased if one is willing to accept an increase in the probability of a bit being mis-read x% of the time, which for some applications (like video and audio signal processing) only results in a small loss of precision.
         Thus, the use of a probabilistic chip can decrease your energy requirements (and corresponding operational costs of computing machinery), but cannot improve the processing accuracy of any applications so using it (although it can speed the chips up slightly since lower voltage utilization means they can run faster without overheating). And, at least for now, one will get (considerably) more speed from parallel processing.

More Proof We’re Overworked … and in Deep Talent Trouble

As per this recent article in Industry Week on “building a lean, mean profitable machine”, productivity in manufacturing has jumped by a record 94% during the past two decades while headcount has significantly shrunk. That’s a productivity increase that’s 60% higher than any other U.S. business sector. Since technological innovations have occurred across the board, one cannot attribute all of this productivity increase to new technologies. A good portion of it is due to blood, sweat, and tears — and people working harder and longer than ever before. And while it looks good on the books, it’s not sustainable over the long term.

There’s a reason that 9 out of 10 employees are looking for a new job. They feel like they’ve been worked to death and that there couldn’t be a job that is possibly worse than their current job. Considering that production is becoming more and more specialized and reliant on precision machinery and technology, this is not a good thing as the industry as a whole is facing a dire shortage of skilled production workers, scientists and engineers. Add this to the predicted shortage of up to 100,000 logistics workers by mid-decade, and you see a deep talent shortage looming across the supply chain.

How does your organization plan to manage it … before it’s too late?

If a Deal Is Too Good To Be True, IT IS!

This is just as true in technology and services as it is in products. If you get four bids for a new technology platform and / or (integrated) services package and three are plus or minus 20% and one is 1/3 of the price, I guarantee that lowball bid is too good to be true. And if you did your homework, you’d instantly know it and disqualify it.

You buy a product or service because it’s cheaper to buy than to build or perform it in house. However, that product or service still has a cost to the vendor, in terms of manpower and resources — costs the vendor has to meet in order to deliver you a quality product or service. If the vendor doesn’t cover these costs, and make a fair profit, one of two things is going to happen — the vendor is going to go out of business trying to serve you at an unsustainable level or the vendor is going to deliver a significantly inferior product or service to stay afloat.

I’m reminding you of this because a number of companies have not only been looking for new solutions now that we’re into a slow recovery, but because a number of companies, desperate to reduce costs, have been rebidding everything under the organizational umbrella, including the supply management platform(s) and service contracts. And in doing so, many of them have been getting unbelievably low bids from a handful of vendors who are desperate to win (new) market share — and the companies are seriously considering these bids. These bids are unbelievable for a reason — they’re not real. They’re up front costs, and as soon as you sign on the dotted line, you’re going to be hit with “change fees”, “service costs”, “upgrade fees”, etc. if you want the same level of service being offered by the competition, who are all in the same ballpark at sustainable bids. Or, even worse, the vendor is just going to give you the platform or an initial spending report, and then disappear until renewal time because the cost only covers platform support, not project or customer support. Or, and this is the worst situation of all, the vendor is trying to build a new business (in a new vertical) and thinks it can use you as a marquis customer to attract new customers, who it will overcharge to make up for the loss on you. If it works, you’re in luck, but the vast majority of the time what happens is that either the vendor fails to deliver, because they didn’t understand the true success requirements or they didn’t understand how much it would cost and how long it would take to make you a success, and then shuts down the business. If you’re lucky, they just shut down the vertical and you get to keep using the platform until you can find a new vendor. If you’re, not, the whole vendor goes tits up and you’re left holding the empty bag.

The worst part is that every month, if not every week, I hear of yet another company who signs on the dotted line with one of these vendors offering “unbelievable” deals that “can’t be matched” — and, even worse, the company is one that should know better (because there are success stories that illustrate it understands many of the precepts of good supply management). Especially when it’s so easy-peasy to determine if a bid is reasonable or not.

It’s easy to determine a reasonable range for a (bundled) technology platform (and /) or service. All you have to do is build a should cost model. Let’s say you’re buying a SaaS e-Procurement platform and want regular project management support, best-practice training, and custom integration to your in-house technology platform. Then you know the vendor will have, at least, the following costs:

  • Platform Delivery & Maintenance
  • Account & Project Management Personnel
  • Development Personnel

If the SaaS license will require 1/50th of their data centre resources, then the base overhead to support you will be 1/50th of their data centre and support team costs. If you require about 20 hours a week of account and project management support and training, then you will require half of a senior resource who has expertise in your industry and categories. If the custom integration is expected to take two man years, than you will need the equivalent of two developers on the vendor’s staff dedicated to you.

Now, if the average cost to maintain a small data centre, or rent part of a data centre, that will support 50 similar-sized enterprise clients is 3M, then you can quickly estimate that it will cost the vendor 60K (+- 10K for a margin of error) just to have you on the books, before it lifts a finger. If the senior resource required to support you on your projects is a 120K to 150K resource, then it will cost the vendor 60K to 75K to dedicate this resource to you half of the time. And if the average developer with the necessary skills is going for 70K to 90K, that’s another 140K to 180K that the vendor needs to outlay to support you. Then, there’s the vendor’s cost of sale, which, depending on commissions structures and expenses, is probably in the 15% to 25% range, and the need for the vendor to make a fair profit, say 10% to 15%, to keep investors happy. If you add it all up, you get:

Cost $ Range
Platform Delivery & Maintenance 050K to 070K
Account & Project Management Personnel 060K to 075K
Development Personnel 140K to 180K
Subtotal 250K to 325K
Cost of Sale 040K to 070K
Profit 025K to 050K
Total 315K to 445K

This tells you that any bids you get in and around the 315K to 445K range are reasonable, that if you get any bids that are more than 600K, the vendor either doesn’t understand what you want or is trying to rip you off (up front), and that if you get any bids less than 250K, either the vendor is planning to not support you to the level you need to be supported, the vendor is planning to make it up later with “change fees” and “service fees” when you’re locked in to a long term contract and held captive, or the vendor is looking to make a poster child out of you and take unfair advantage of the relationship (and then leave you holding the empty bag if things go south).

Regardless of why the vendor gave you the unbelievable bid, one thing is clear. If you accept it, you will get screwed.

Oh No! App Mania Has Hit The Supply Chain!

Prepare to be nickel-and-dimed one feature at a time. It’s coming.

Unlike the author of a recent post in the SCMR blogs, I am not pleasantly surprised by the amount of effort being put into smartphone app development right now. First of all, there’s only so much you can do on a smartphone screen. For the vast majority of analytical supply management applications, smartphones just don’t make sense. Secondly, unless you have constant access to free wifi, you’re not going to want to download megabytes of data and push your mobility bill through the roof. Thirdly, the real value in supply management systems materializes when data flows from one to the next — and all of the relevant data is available. Data stuck in a standalone app that doesn’t do much more than display that data isn’t very useful.

Plus, do you really need 100+ WMS apps? Really? No! There aren’t 100 WMS applications, and if you think there are, then what you are really buying is 100 different instances of a WMS workflow engine custom tailored to a specific situation. You’re paying for 90% of the same functionality 100 times over. I guess it’s good for the vendor if they can create 100 apps by only doing the work required to create 11 and charging you for each and every app, but it’s not good for you.

HighJump, one company that plans to release up to 100 apps a year, may try to tell you that it’s a great idea because it is a way to get new functionality in your system without waiting for a new release, but you don’t need apps for that. If you’re using a true multi-tenant SaaS application, then you get every update the vendor makes as soon as its available, and you get it all for one maintenance fee.

In other words, if you want to needlessly empty the corporate bank account, there’s an app for that.