Data processing, 1959. Image by James Vaughan, CC license

If you search “technologies for supply chain evolution” you will get 17.8 million results, many of which will mention IoT, robots, driverless vehicles, mobile technology, predictive analytics, network optimization software, and 3-D printing.

The presumption is that industry will want these solutions because of the coming added complexity and demands of the modern supply chain in the digital age.

But most supply chains still confront very basic problems and inefficiencies; they haven’t in many cases even fully deployed yesterday’s technology, so delivering product with drones doesn’t make the priority list.

So which technologies are key to the supply chain of the future? The same ones that were introduced 15 years ago that companies still haven’t adopted! This is unfortunate, because many of these technologies were designed to automate the supply chain office, and without automating the supply chain office, all of that promising talent you are committed to developing is typing away at tedium.

Just visit your customer service department, where very smart and capable humans – some with $120,000 college degrees — are reading data from pieces of paper and entering them into systems, an act which, according to supply chain technology gurus, should have disappeared years ago with the introduction of EDI, integrated systems, and OCR.

Supply chain managers are managing minutiae, not the supply chain.

Where does the minutiae come from? From the 18,000 to 30,000 phone calls or emails and the 120,000 to 350,000 manual interventions a typical ($2 billion +) company must make just to ensure proper system processing of its order-to-cash flow. And this is with all the normal ERP and associated technologies, such as EDI, that a company of that size normally uses.

Companies still have teams of people shuttling information and data all over the supply chain office, and because they are busy shuttling information they don’t have time to look at it to make sense of it and use it to improve your business.

Automating the supply chain office is one of the cheaper technology moves you could make. The EDI and OCR of the 1980s has gotten better and easier to deploy, and when strung together with complimenting technologies, can automate nearly all of your order processing, exceptions management, invoice discrepancies and customer claims work.

Automation also has the enormous benefit of – by definition – digitizing every piece of data from every transaction. If all of your shipping and invoice claims are scanned, stored, and sent to transactional systems for disposition, all that data is available to study for patterns, relationships, and other insights that can mean immediate savings. Like getting a view of the forest.

Want to evolve? Make the supply chain office run itself. For a $2 billion company selling to major retail outlets in the US, this means zero human intervention for every one of 60,000 deliveries a company of that size is likely to make. Most business rules are simple to automate: if X order received, send to Y location and reply Z back to customer.

A truly evolved supply chain office is one where all human assets are users of data, not movers of data, creators of opportunity based on a view of the forest, and customer relationship builders.

There’s the human factor to consider.  Photo: Workers at the Jamaica Plain Post Office, 71 Green Street, Boston, 1920.  Boston Public Library, CC license

Have you ever spent, say, half a million dollars and nine months of effort with new software or changes to existing systems for purposes of automating processes, only to discover that afterwards the processes are just as inefficient as they were when you started?

I’ve worked on several of these projects.  People claim they can save hundreds of people-hours per week if only …. Usually these projects require custom development of existing software or additional software to eliminate process steps.  Why do these efforts fail to deliver promised benefits?

A ‘soft’ or hard-to-prove ROI .  Frequently labor savings from automation do not result in actual reductions in labor cost; instead the argument is made that the work force can engage in more “value-added” work.  Consequently, if you had 30 people processing customer orders before the automation project and 30 people after, it’s still possible to call the project a success, because you can say that now the customer service team is pursuing more “value-added” work.  But is it?How do you know?

A tendency to stick to comfortable work routines .  When people have been using the same systems to do their work for 10 years, it’s very hard to change their routines and not to do certain steps inside those systems because the steps are no longer necessary.  People are going to do what is comfortable for them, especially when they see only a marginal benefit to changing.  You need a big change management campaign to ensure new more efficient practices are followed.

Higher than expected  costs to modify software .  It is also a human tendency to underestimate costs of making changes to existing software.  In the case of ERP systems this cost can be astronomical – not so much because a lot of hours are needed to modify the software – but because the original code was never intended to be changed in the way we want it to change.  You can only do so much change before the entire program has to be overhauled.  Everything inside a complex software program is interconnected.

So what do you do? 

Avoid big and expensive modifications to software as a way to automate business processes.  The people who created the software had a specific purpose in mind; that purpose may or may not be compatible with what you imagine to be a more efficient business process.

If you want true automation, outsource the process altogether .  Taking pieces of a process and eliminating them through software modifications might be minimally helpful, but it’s not a big enough change to modify behavior.  Yes, to make a process more efficient you may have to remove it from those people currently performing it.  The adage about teaching an old dog new tricks is true.

Outsourcing the process does not mean you can – or want to — eliminate whole departments.  But it does ensure that certain activities – non value-added processes – no longer exist within your organization, for anyone to waste any more time on.

If you visit OmPrompt in the UK you can stay in the nearby dreamy, historic, and beautiful town of Oxford. Photo by Tejvan Pettinger, CC license.

What if you want to make sure all of the data on a customer’s order is correct before it is processed in your systems?  Or if you want to stop orders that you can’t fill based on time constraints?  Many firms at first apply humans to these tasks (customer service), but what if you process 55,000 orders per year?  There must be some way to automate this screening process, right?

There are at least three ways to handle this problem with software.

The first is to modify your ERP system, to incorporate the rules you will apply to every order that is received.  Not only will the ERP system apply the rules and identify exceptions but alert you to those orders that must be stopped and changed before proceeding through the order cycle.

This is certainly a viable option, but you’ll find yourself pouring lots of money into it, with no end in sight, because as soon as you successfully incorporate a set of order checks, the business will come up with a new set of checks based on a new logistics or pricing strategy.  Most ERP systems are not built with user-configurable parameters; changes require programming by expensive resources.

The second is to build some type of middleware filter.  This is a custom application that every order would pass through before entering your ERP system.  The middleware would catch and hold up non-compliant orders.  A lot of firms adopt this route, because it is so much less expensive and much more flexible that tearing into the guts of an expensive ERP system.  And there is no shortage of eager programmers, within company IT departments, who want to build such a tool, in the language they know best on the computing platform they know best.

This second option is better than the first, but barely.  The second option makes you dependent on the genius that builds and maintains that middleware tool, and that is almost as bad as being dependent on an ABAP (SAP code) programmer.  And there will be instances where no matter how brilliant your middleware architect, certain business requirements will be beyond the capabilities of him/her and/or the software they have created.

The third option is to outsource your order capture, partially or entirely, to a firm that has the systems agile enough to handle the ever changing order filters your business may want to apply.  The advantage of these firms is that they serve as adapters: the incoming order is converted to a standard computer script where all of your rules can be applied; the script is then converted to the digital code (such as EDI) needed by your ERP system.

One firm that does this is UK-based OmPrompt, which specializes in automating business processes for large firms all around the world.  Their services are based on transaction volume and are very inexpensive relative to software modifications or building your own system.

I favor outsourcing the job not so much because someone else can do it better (although that is true) but because you cannot predict what is coming down the road in terms of new business requirements.  You need an open system, an infinitely-adjustable capability, if that is possible.  And firms that make their living on being adaptable to requirements like yours will always have more value to offer.