Check out the design for the OneDown mousetrap, a Gold Award Winner at the recent International Design Excellence Awards that was created by the DSK ISD International School of Design in India.
Monthly Archives: July 2010
SI’s Policy: No Time Limit and No Fixed Schedule on Briefs
One week ago today, Spend Matters (“SM”) announced that it was limiting briefings to
a one-hour-and-twenty-minute interval on Thursdays and a
two-hour-and-30-minute interval on Fridays. SM stated that
briefings would be limited to 40 minutes, including 10-15 minutes of introduction/background,
the remaining time to be spent on “the presentation or announcement”. SM further stated that it requires
reference contacts for a follow-up call, and that it is limiting follow-up demos to 30 minutes.
While it is not Sourcing Innovation’s place to comment on the policies of other blogs in the space,
the doctor would like to take this opportunity to make SI’s policies, most of which are very different from the above (and laid out in the FAQ), very clear.
- SI will schedule briefs at times of mutual convenience. If this means after hours, that can be arranged.
- Although initial demos are usually limited to 60 or 90 minutes, SI does not impose time limits of any kind on briefs. If a brief needs to be continued,
then it will be, for as long as it takes. SI is interested in your solution. Very interested.
- SI is not interested in “announcements” or “presentations.” Rather, SI wants to see your
solution in action. Plan on a deep dive into your solution, followed
by an intelligently-written post, assuming that the doctor believes that your solution solves a
real problem in our space. Note that the doctor is very respectful of any software solution that
solves a real problem, because he knows what it takes to bring such solutions to market, having
done it himself on multiple occasions.
- SI is generally not interested in making reference calls*, because the doctor doesn’t care about your Marketing department’s ability to find customers who think your solution is the greatest thing since sliced bread. He is interested in your solution, not third-party opinions about your solution. This is not People Magazine or Gartner, so the opinion of a random “celebrity” customer, regardless of how prestigious his or her corporate logo may be, is irrelevant here. Lemmings are plentiful, and they come in all sizes.
Finally, SI sees no issue in trying to “keep up with” vendor products and announcements. Although many vendors make routine “exciting” announcements about new functionality and so forth, in practice these announcements are usually just marketing noise, having little to do with important functionality changes and enhancements. SI is confident of its ability to cover real enhancements and real innovations when they really occur, and can see no immediate danger of running out of cycles to do so.
For vendors who require a Non-Disclosure Agreement (NDA) before providing a demo, SI is not interested in reviewing your solutions. the doctor’s assumption is that organizations requiring an NDA before doing a deep dive for a reviewer are irrationally hiding functionality that is released and in the public domain anyway, so their motives for trying to conceal that functionality are suspect.
Thank you for your attention.
* The exception being when you make a strong claim about savings, ROI, etc.
A Hitchhiker’s Guide to e-Procurement: Purchase Orders, Part II
Mostly Harmless, Part VII
In the last post, the purchase order was defined as well as some of the requirements for its generation. This post will address the challenges associated with purchase order generation, some associated best practices, and the benefits that could be expected from an appropriate e-Procurement solution.
Common Challenges
- Requisition Partitioning
A requisition contains requests for multiple goods or services, which are covered by multiple contracts at multiple rates depending on SKU, volume, or other terms. Which contract? Which rate? Which terms?
- Forward Matching
How will the purchase order be matched to incoming goods receipts and invoices?
- Duplicate Detection
How does one detect if multiple purchase orders contain a requisition for the same good or service? How does one detect if duplicate purchase orders were accidentally cut?
Best Practices
- Automatic Generation
The system should automatically generate the necessary purchase orders from approved requisitions.
- Automatic Price Confirmation
The system should automatically verify that contract or catalog prices are being adhered to.
- Automatic Distribution
An approved purchase order that sits on someone’s desk waiting to be sent can hold up the business or a production line if the parts or services are not delivered on time because the supplier(s) did not get the purchase order on time. Once a requisition is approved, the purchase order should be sent automatically
Potential Benefits
- Reduced Lag Time
An e-Procurement system can automatically create and distribute purchase orders as soon as the requisitions are approved.
- Reduced Overspending
The system can automatically grab and populate the purchase orders with contract pricing. Some categories, like office supplies or electronics, see a lot of overspending because buyers requisition at catalog, but not contracted, rates or don’t buy in the appropriate quantities (which can be flagged and corrected during the approval process).
- Reduced Errors
The system can automatically pull up the right codes, the right templates, and the right prices so that the supplier isn’t sending it back with a request for further explanation, which would only delay the process further.
Once the purchase orders are distributed, the next step is to wait for delivery and issue the goods receipt, which is the subject of the next post.
Next Post: Goods Receipts, Part I
China’s New Labour Militancy
Editor’s Note: Today’s post is from Dick Locke, Sourcing Innovation’s resident expert on International Sourcing and Procurement. (His previous guest posts are still archived.)
In my last post I wrote about the loosening of the controls on the value of the yuan. Since then, the yuan has weakened about 1%. That’s one of two developments affecting China sourcing.
The other, which I think will have more immediate effect, is new labor militancy among the employees of export related industries. They are realizing they are in a very strong bargaining position, because they are working for first tier suppliers of consumer products, such as computers, telephones and cars. The Apples, Dells, HPs and Hondas of the world aren’t going to stay customers of companies who take (with or without government involvement) repressive measures against labor militants.
This is unlike the situation in Mexico, where the federales broke a miner’s strike a few weeks ago. Miners are about as far back up the supply chain as you can get, so it’s unlikely the mine’s customers are going to feel consumer pressures. The mine is in Cananea, and is more or less an historic site because there was another famous government-assisted strike breaking there about 100 years ago. In that 1906 incident, the US Army got involved, 60 miles inside Mexico.
In China, the suppliers are generally conceding to labor’s demands. Will that lead to a massive departure from China? Not in itself, particularly in the electronics industry. As a rule of thumb, electronics assembly costs are 80% material, 15% overhead and only 5% labor. A small increase in labor costs at the assembly level is unlikely to be a deal-killer. That’s helpful, because resourcing all that electronic assembly work out of China is going to take years. Foxconn alone has 800,000 employees in China. Resource that!
Thanks, Dick.
Tips for Evading the Black Swan
The black swan has been on a rampage in recent times, taking down supply chains, companies, states, and even countries in his wrath, which seems to have no end. It’s foolish to assume that he’s not coming for you, because, even if you’re not on his list, it’s only a matter of time before you are. He’s as determined as robot santa claus, and just as indestructible. But if you’re ready for him, even if you can’t stop him, you can survive the encounter, and with the right blast shield, even minimize the damage. So how do yo do this?
Nassim Nicholas Taleb’s exceptional article on The Fourth Quadrant: A Map of the Limits of Statistics has some great advice for minimizing your chances of encountering the black swan, and even minimizing the damage if you can’t avoid him. And to make sure there’s no misunderstanding, I will use plain English, and not statistics (which most people, including the “experts”, don’t really understand), in my presentation of these tips.
- Redundancy
You can over-optimize and over consolidate. You need multiple sources of supply, multiple products, and multiple channels.
- Avoid the Long Tail
Yes you can make money in the long tail, if you’re lucky, but the further you are from the norm, the harder it is to predict what will work.
- Don’t try to Numerically Model Atypical Events
You can’t predict future stock-outs based on past stock-outs or the degree of future demand surges based on historical demand surges. They could be the same, or be off by a factor of 10. That’s why they are atypical. Address them generally, and you’ll be better off.
- Take Your Time
Only time can reveal the true nature of a cycle. Depending on what you’re trying to model, that could be months, years, or even decades. If you avoid drawing conclusions too early, you’ll be better off.
- Don’t Reward Luck
Just because someone made a foolish bet and won doesn’t mean they should be rewarded. The more extreme the bet, the more likely you are to lose. Don’t encourage ridiculous behaviour.
- Don’t Measure What Can’t Be Managed
For example, the “average time” between stock-outs or demand surges is meaningless, and it will just increase the desire for your team to “model” the situation, which will give you the illusion that you understand something you don’t, and that you don’t need to have contingency plans for “unexpected” stock-outs or demand surges because you modelled them.
- What’s the Nature and Magnitude of the Uncertainty?
In NPI, the uncertainty is that the team might fail given the resources assigned to it. The nature of the uncertainty is positive (if they succeed, you win) and the magnitude is limited to the investment. But in chemical processing, the uncertainty is that a storage tank could rupture, contaminating the local environment. The nature of the uncertainty is negative (if it the tank ruptures, the environment gets damaged to some level) and the magnitude is large (if the chemicals reach a lake or the groundwater table, the local population is screwed). Put your efforts on creating emergency plans for large negative uncertainties first, as those are the events that can bankrupt the business.
- Do Not Confuse Absence of Volatility with Absence of Risks
For example, if you look at the graph of daily variations in a derivatives portfolio exposed to U.K. interest rates between 1988 and 2008, almost 99% of the variation occurs on 1 single day — when the EMS (European Monetary System) collapsed. On almost every other day, variation was less than 1/100th of a percent. This is not dissimilar to the eruption pattern of Mount Vesuvius (which buried Pompei and Herculaneum in 79 AD). If you plot a daily graph, it’s typically flat for 30 to 50 years, until one day a massive eruption wipes out the local area. Remember, the black swan will show up where you least expect him.
- Most Risk Probabilities are Lies
A rare event that happens once every 30 years does not have a 3% chance of occurring every year. The chance is typically dependent upon whether or not there is a confluence of initiating events and factors, and could be 0.03% or 99.3%, depending. Furthermore, the presentation of a risk statistic has a significant effect on it’s impact. People are unlikely to heed a warning for anything that only has a 3% chance of occurrence, but very likely to at least give serious thought to any event that will happen once every thirty years with 99% certainty.