Courtesy of Brian Tinham of Works Management

Selecting new ERP software is only part of the task; successful system implementations also need re-optimised business processes and workflows.

When the time comes for changing or upgrading ERP systems, most manufacturers that have walked the walk will agree, retrospectively, that selecting the software itself should have been a relatively small part of the story. More often than not, it wasn't - and that mistake is probably the single biggest reason for so many projects failing to deliver fully against their business justifications. 

Talk to almost any clued-up project manager and you'll hear the same advice. Don't procrastinate over your choice of ERP: you're never going to find a system that matches precisely what you do. And, if you pursue that path, you'll be looking at high levels of customisation, high implementation costs and years of delay and regret. 

Instead, spend more of the valuable time donated by your cross-functional project team on what matters to the business. Common consent suggests taking it from the top, meaning undertaking the SWOT analysis and using that to define the business strategy. Then you're into understanding what it's going to take to align the IT to deliver on that - taking into account any new capabilities. 

And not just the IT: given that all modern systems are effectively enablers, plan to use the 'opportunity' that changing ERP brings to re-optimise your business processes. That's key to realising value, whether in terms of increased business agility, better customer service, slicker supply chains, reduced admin or all of the above. 

Of course, due diligence remains important. And, yes, that means selecting your IT (and consulting) partner, and rediscovering and documenting the current state business processes and IT involvement - from top floor to shopfloor. It also means working on business-wide data cleaning (customers, suppliers, parts nomenclature, BoMs, routings etc), as well as considering the migration method and the rest. But, without wishing to labour the point, your peers will tell you all that needs to be viewed as part (albeit an essential part) of the journey; not the reason for it. 

That's the top level thinking. Meanwhile, at the sharp end those embarking on this journey invariably state that getting their heads round the current and future states - and, in particular, mapping business processes and visualising where IT fits in - is difficult. For most manufacturers, the complexity is massive and the capacity for any new IT to help change the way things are done is probably unknown - and also slightly scary. There are also always those who claim to believe that the way things run right now is pretty good, that the law of diminishing returns pertains, and that 'value engineering' is just another meaningless buzz phrase. 

So best advice might be to grade potential software suppliers less in terms of the software they offer - although there remain differences and no one is saying this demands zero attention - and more their consultancy services, and relevance to your operations and sector. That said, also look for IT partners that can specifically help with business process discovery and re-optimisation, both in terms of methodologies and graphical visualisation software. 

Software assistance 
ERP companies have offered the latter facilities for years, with varying degrees of success. Who remembers, for example, Geac's System21 Aurora business process modeller a decade ago? And it's also the case that so-called 'pure play' BPM (business process modelling) software companies, such as Tibco, have made, and continue to make, a good living out of their offerings - albeit mostly from the big corporates that can afford bigger ticket projects. 

Whatever your perceptions though, some modern BPM systems are surprisingly helpful, enlightening and easy to use. Most now also help to define system functionality, workflows, screens, data handoffs etc, which in turn means serious potential for faster deployment. Take SAP, for example. The software giant is by no means alone, but its Netweaver BPM suite, although embryonic in terms of uptake anywhere outside the system architecture specialists, does look impressive. 

Adrian Simpson, SAP's head of technology, makes the point that most manufacturers' business processes don't start and end in their ERP software. "So project teams need to be able to understand, configure, prototype, validate and then execute business processes, no matter where they go. That's why we introduced Netweaver BPM, which is aimed at helping them to customise business processes [not software] that go across applications, and to extend them into what we call 'next practice' - in other words, not limiting them to today's best practice." 

For SAP, this is seen as an important differentiator, since it means that project teams can not only visualise default processes, but also configure improvements specific to them that deliver competitive advantage, without bespoking the code or getting bogged down in the detail. "It allows them to model what the business processes looks like, using an architectural 'swim lane' view, with systems on the 'x' axis, and actions, comments and functions on the 'y' axis, while the connections across and down represent the process flows," explains Simpson. 

"Considering suppliers, for example, the screen might open up processes for bidding and taking quotations through to contracts, then into procurement and so on to manufacturing," he continues. "So you're mapping across all relevant systems, and with all the orchestration revealed in the schematic right in front of you." 

And with that level of logic exposed, it's not difficult to see that business analysts can then go on to the next level - designing the look and feel of operational screens, with users looking over their shoulders and making amendments in real time. "It's easy to do the prototyping, and it's iterative. Users can say that's not quite what they want, or suggest changing part of the flow, and the system can be rebuilt on the fly," enthuses Simpson. "Not only that, but at another level, once all the information and metadata is complete, along with the integration requirements and service calls to relevant applications, we then have something that is executable. We hit the button and that design becomes something that can actually run." 

All well and good, but what if your system isn't one of the latest SAP offerings? Additionally, suppose documentation is poor - certainly at the business process level. Netweaver BPM isn't going to help then, is it? Just so: and that observation would be just as valid for any other mainline system and its associated business process mapping suite. 

Suffice to say that SAP's answer then is its Solution Manager, which, in its case, integrates with Aris' independent business process mapping software. Simpson explains that the former enables reverse engineering of the embedded business processes as originally implemented. Then the latter 'componentises' what's there and helps users to make improvements, effectively re-using the original software. "We have customers happy to run the tools and do it themselves, while others use the services of our system integrators," he says. 

So much for the theory; what about practice? John Routledge, IT manager at niche engineering equipment manufacturer Drallim Industries, is one who believes fervently in the power of business process optimisation. He says his company's own ERP upgrade, which went live a little over six months ago, is testament to the value of embracing change and recognising that so-called best practices might actually be just that - and probably therefore better than your own. 

4,500 hours saving 
He points to top line improvements as centring initially on administrative efficiencies that are saving Drallim 4,500-plus hours per year across 37 users. Admittedly, that was largely as a result of switching new group acquisition, Horsell Electronics - which had hitherto been underserved by IT - onto its new templated ERP. Nevertheless, that change, he says, led to reduced lead times in all areas of the business, including report preparation, serial number control, goods receiving, supplier payments, general ledger management and systems administration. Cash flow was also significantly improved, through an electronic accounts receivable payment interface that has seen payment terms drop from 60 days to an average of just five. 

In brief detail, the Hastings-based firm, which specialises in equipment for the aerospace, industrial and utilities markets, upgraded from an Avante ERP system to Epicor's latest 'next generation' suite. Routledge concedes that, like so many others, his firm had moved on significantly since the Avante deployment in 1999, but had not maintained its documentation. So for him, that was the first job. 

"We formed an ERP project team of senior managers from all departments, selected for their knowledge of company processes, but also their authority to make changes, and we started by re-documenting everything we do. It took around three months, getting down to action-level detail, whether on the computer system or not. But, armed with that, we felt prepared to look for improvements - which we kicked off by playing with Epicor to see what it could do." 

Once the team had a reasonable grasp of the system, Routledge says the next task was modelling the compete quote-to-cash scenario. "We tried to map that to our as-was processes and, where there were variances, we worked on making changes to suit the new system. If you like, business process optimisation was almost forced on us by our insistence on 'vanilla' ERP. Sticking to the Drallim way would have been a very arrogant approach." 

Routledge observes that anyone doing this needs to take a structured approach and tackle it in bite-size chunks. Drallim went for a simple flowchart, with action boxes, coloured to indicate whether or not they were inside the system. "With processes printed out, the team could challenge one another on why certain things were being done and what value they were bringing. It was all about sanitising processes and, nine times out of 10, the system preferences represented improvements." 

That second phase only took four to six weeks, he says, because, by then, the team knew what they were looking for. That's an important point: executed methodically, business process optimisation needn't be a nightmare. For Routledge, however, another part of making change work was nothing more sophisticated than user familiarity - and one way of encouraging that, he says, was to run data cleaning and migration in parallel, and to start on day one, rather than wait for mapping to complete. 

"By using our data early, as part of the migration and cleaning processes, people saw that the new system was delivering the right information. They could test MRP, for example, in terms of lead times, parts being bought from the right suppliers, at the right price, etc - and all the time they were learning. Also, they were motivated to look at fields, not just in terms of consistency, but also reconsidering groups and classifications so that, for instance, sales and production could filter and sort on the same data." 

It worked, with the approach expanding from the cross-functional team to users, building in the usual top-down training, but also with testing and development in parallel. And the results speak for themselves: improvements on go-live, as above - driven by everything from slick screen navigation to efficient analytics and workflows. 

Routledge's advice: business process optimisation is not the only aspect you need to get right, but it sure is right up there. "Any upgrade should be treated as an opportunity to reinvent your business processes. Manufacturers rarely have time to rethink what they're doing. But if you're putting in a new system anyway, why wouldn't you work it to add value?"