Technology Tools

Quentin Samelson, Nokia supply management exec and Motorola veteran,  says that technology tools hold enormous opportunity for industry. In supply management, Samelson sees a second round of interest in optimization, this time as a tool that can bring real benefits. Optimization isn’t really a new concept – it’s been around more than a decade. But early expectations outstripped capabilities, so few companies used the technology effectively. The perception in the early 2000s was that all we had to do was feed in the data (often outdated, incomplete, and relatively low-quality), turn the crank, and out popped the optimal plan. Well, it didn’t work that way, and it still doesn’t – but as we approach the middle of this decade, companies are much better prepared to (1) provide the detailed, high-quality data needed, (2) provide that data on a regular, timely basis, and (3) act on results.

Further, some companies look at the concept from the standpoint of “help us make a decision about scheduling this manufacturing line,” or “make the trade-off between building more of product A vs product B, ” instead of trying to optimize everything all at once. Realistic expectations combined with better data mean that the technology is finding its niche.

Optimization provides a set of tools to help planners decide how to best juggle inventory, supplier risk factors, and market demand over multiple global sites. In telecommunications, where for instance customer demand can shift dramatically with little or no notice, and companies must balance product volume across multiple facilities simultaneously as they expedite (and de-expedite) suppliers, optimization software helps rebalance product volume against material orders and plant volume. Keeping all the value streams balanced is a complex task that requires really good and visually simple technology assists. 

Unfortunately, says Samelson,  there are still many companies that don't even TRY optimization; some that try it often don't install it or they lose interest after initial efforts. But all this is changing based on these three factors that I’ve observed over the past 3 – 5 years:

1. ) The technology has become more user-friendly. Where we once were forced to rely on multiple data sources – screens, print-outs, dashboards – the user now has data and exception reports easily and instantaneously generated on-demand. Selecting a range of products to analyze, pulling in current MRP data, and running the calculations to simulate a change in manufacturing schedule or product mix can be done far more easily than before;

2.)   The amount of data that companies have available for analysis has finally reached critical mass. Not every company can provide a comprehensive daily exception report, but most companies can provide   users with daily updates on at least their most critical situations. The days when everyone waited for weekly MRP print-outs – then pored over them for a week – are over. More analysts rely on dashboards and “control towers” for daily updates – in some cases they can drill right into the dashboard to examine precipitating factors;

3.) A whole generation of production & inventory control managers believed that you could improve factory/warehouse utilization, improve customer service or reduce inventory, one variable factor at a time, and never all three at once!  In the first decade of the 2000s most of us learned that that wasn’t true – we were in fact required to improve all three at the same time. But we’re not done yet. There will be continued pressure to reduce inventories while simultaneously improving service levels, and this will create a need for new tools.  We’ve always known that the old inventory formulas didn’t connect well with detailed service level requirements over global networks, but approximation and painful experience were the only aids we had to improve performance, and we often settled for ugly compromises.  But new powerful optimization tools let us get very specific about customer service requirements – if we have one high-value customer that requires 24-hour service, we can set the dial to trigger the calculation that will do it.  If another market segment can work with 48 – 72 hour deliveries, the reset calculation for this different service level is immediate, and we’ll know the cost difference right away.   

Behind all the logistics and supply challenges, the financial implications remain.  Samelson doesn't see any reduction in the need for companies to focus on managing costs and inventories while improving customer service (see  3. above!). Economic predictions for the next couple of years suggest that there won't be any magic new technologies or business models that will eliminate the need to keep working hard on efficiency and effectiveness!

What will blindside companies?  Samelson sees weak spots crying out for attention at the points where companies reduced costs through outsourcing and lay-offs. Of course, a few organizations replaced weak internal functions with stronger external services, but more often they lost experience-based knowledge and skills, and these can’t always be replaced by outside services, especially when those services are off-shore.  Companies are learning that replacing internal resources with external services CAN be done with quality -- but it's also easy to get it wrong - and what muscle remains after strategic outsourcing needs to be rebuilt and supplemented with newer tools.  

***

AND A FOLLOW-UP QUESTION FROM SUBHAJIT G. VIA LINKEDIN:

And Quentin Samelson's Response:

Hi Patricia,

This is a HUGE question to answer because it's so open-ended. Here's a way to break it down:

1. For some parts of the supply chain, there are dedicated, specific tools available. A good example is inventory optimization for sales items. If you type "inventory optimization tool" into Google, you'll find lots of companies that provide this sort of tool.  (However, many of these tools are not as effective at optimizing at the levels below the sales items.) 

Another area where optimization technology has been applied effectively is in internet negotiations. Companies like Emptoris (now part of IBM) and CombineNet have built optimization into their tools, so that users can make the most effective sourcing decisions once their negotiations are complete.

2. For "supply chain optimization" there are separate tools -- such as Jonova. These tools look more at the value chain itself, and help users identify "more optimal" solutions. These are especially useful in improving the logistics of moving raw materials to processing facilities, then moving product to customers.

3. As your contact seemed to imply, there rarely is an optimization solution that can be used throughout an organization... except -- the underlying "engine" that powers most optimization tools is a mathematical program called a "solver." These can -- theoretically -- be applied in different ways in different departments. IF your company is large enough to hire a mathematician specializing in this technology, you could theoretically apply this tool throughout your company. 

(This might sound far fetched, but there are companies that do this sort of thing every day... I just don't know of any manufacturing companies that do! The companies I'm thinking of are in the financial services industry. At least one company I know of has hired masses of  PhD mathematicians to figure out better ways to optimize stock trading.)

4. If there isn't a commercial tool available for the challenge that you're facing, and you can't hire a mathematician -- then you have essentially two choices: 

a) hire a consultant

b) build your own

This sounds challenging (and it can be!), but the the plain truth (in my opinion) is that we often have unrealistic expectations of what "optimization" means. People get the idea that they can just buy a tool, hook it up to their ERP system, and press "optimize." That's rarely (if ever) true. We often don't have enough data to truly find that perfect, absolutely optimal solution. What optimization does is help you find the best available solution. The good news is that, for smaller situations, often you can find the best available solution using tools that are already built into Excel. The bad news is that you need to take a  scientific approach to the problem. That means collecting the necessary data, ensuring that it is complete and of sufficient quality, and getting agreement on what an "optimal" solution would be.

The classic situation that confronts supply chain managers is one where there is a shortage of a critical component. It may be used by multiple facilities at the same company -- all fighting to get enough of that component to support their customers. Solving this problem, more than anything else, requires agreeing on what an optimal solution would be. Is it maximizing profits for the company? Or minimizing customer dissatisfaction? Or focusing on strategic accounts? Once you agree on that, the rest is mostly math.

Regards,

Quentin

***