Your guide to choosing and implementing ERP

manufacturing business strategy system selection Your guide to choosing and implementing ERP This may be revision, but for most users that don’t do it every day Brian Tinham provides the collected ‘dos and don’ts’ for shortlisting and implementing ERP systems efficiently and well obody looks forward to a serious business IT project, and that applies whether you’re installing a new ERP system (or similar), or ‘simply’ upgrading or changing an existing one. It’s hard work. It comes on top of the already demanding day job. Yet the success of your company depends upon it. And not just now but for many years to come. What’s more, everyone has heard a horror story or two from someone. No surprise then that many are cynical as well as daunted. But it doesn’t have to be like that. Times, systems, functionality and, most important, methodologies and support, have all moved on immensely in recent years. And so has the body of understanding of hype, blind alleys and on the other hand, opportunities and best practice. So it is now perfectly possible to turn what might well once have been an unrewarding slog (likely to end in failure) into a structured project with manageable milestones and relatively quick wins for the business along the way. Here’s how to do it right, keep your workforce sane and on your side, and save serious time and money into the bargain. Do this and you’ll also get as rapidly as possible to a project that will be judged a success not just at the time but into the future and business-wide. n not just the IT folk, at least not at the beginning. Strategic business thinking and vision need to come first. And following that you need excellence in management leadership, communication, education, training and change management. You also need to understand the scale of workforce commitment required, and make proper provision for day jobs in order to keep the existing business alive. Ignore any of this and you won’t just get a system that fails to meet expectations. You’ll also miss opportunities, see costs and timescales escalate, disappoint your stakeholders and have every cynic on site your saying ‘I told you so’. Secondly, while on the subject of time, don’t procrastinate or let the jobsworths get in the way. The IT world is littered with organisations that took ‘due diligence’ way beyond what’s reasonable. That’s serious waste of business time and effort. It’s also an unforgivable opportunity cost – management time and money wasted as well as business benefits and competitive advantage damagingly delayed. People and policy TOP 10 MANAGEMENT TIPS: I Make management commitment clear and real I The clock is ticking, so quickly identify key improvement/change goals and priorities consistent with senior management vision for the business I Set the project scope and stick to it, consistent with balancing the changing needs of the business I Empower those you trust across the departments and sites to make it happen with diplomacy, certainty and speed I Nail down the phases and change management with staged objectives covering technical requirements, education, training, communication and meaningful KPIs I Project management needs to favour a sequence of quick wins based on firm foundations aimed also at pragmatically achieving the long term goals I Don’t underestimate the data cleaning, standardising and migration effort and timescale – in fact, start all that yesterday I Don’t cut the education or training budgets: you will regret it for a very long time I Be advised: good consultants have valid and useful roles to play that can make a big difference to both the experience and the outcome I Stay on the case Take it from anyone that’s done any of this for a living: good system implementations are about the right people, processes and business culture – not pandering to individual departmental requirements and certainly First and foremost Manufacturing Computer Solutions • Shortlist 2006 • www.mcsolutions.co.uk 9 Step by step shortlisting guide Senior management The first task belongs to senior management. Get your business strategy and vision clear: where do you want this company to be in a few years – and what do you think you need to do to get there. Consider your business drivers and trends; then look at existing business processes, not systems, and see what isn’t working well now and certainly won’t cope in the future. As Ashley Mulhall, manufacturing systems consultant with Allman Horrocks Consulting, puts it: “The management team needs to break the company down into high level components, like manufacturing, admin and distribution, and then look at the business processes under each of those headings to see how well they are performing. They also need to establish the critical success factors going forward. What do they have to be better at than the competition to deliver on their growth plans?” There are tools out there to help. Cranfield School of Management, for example, has a software tool that helps management model areas of weakness. It also offers a database, in association with Gedas, that reveals what other companies similar to you have chosen to do, which can be enlightening. Allman Horrocks’ approach is about constructing a business process matrix to reveal the weakest of the critical success factors and hence the priorities. Says 1 looks to him for the answers. But we want them to take responsibility and ownership.” And he makes the point that, with a senior management sponsor and the requirements documentation in place there should be no danger of straying from the agreed strategy. One key point though: the membership of, and resources granted to, the project team are very important. “Companies almost always fall foul of the project team being asked to do too much – meaning the day job and the project,” explains Mulhall. Even firms that have done it before don’t seem to remember how much effort is required to select and implement an ERP system. “If the day job has to take precedence, the wobbles set in and eventually the real effort falls on the shoulders of just one or two individuals who work themselves to a standstill,” he warns. “Insist that every point is seen and scored in a structured fashion” Ashley Mulhall, Allman Horrocks Consulting Mulhall: “We never ever talk to clients about systems at this stage – it’s way too early. We suggest using the matrix to drive the production of succinct business requirements documentation and thus the functional specification of a system to deliver it.” And he adds: “But at that point we move the debate away from senior management and start to pull in the cross-functional project team.” Consultants generally say they can become part of the solution to the manpower and knowledge issues you will face. They would wouldn’t they? But think about it: using a consultant might look expensive but the payback is likely to be a shorter, more successful project born of solid experience, honed business analysis and project management processes, objectivity – and realistic internal workloads consistent with keeping the business going. Interestingly, none of the consultants I spoke to suggested doing this work for their clients: quite the reverse. Mulhall is typical: “We don’t do it. If you outsource the project to a consultant, what happens when they wander off 18 months later?” The consensus recommendation is something like 50 to 60% responsibility in the shortlisting and selection processes, falling to 20% during implementation itself. And as for being tied to, or at least favouring one ERP system or vendor over another, it depends on who you choose. Allman Horrocks, for example, insists it is independent, claiming to invest significant time and effort in keeping up with system technologies, vendors and their project and industry experience. 3 Use consultants wisely 2 10 Now it’s middle management’s turn. Allman Horrocks recommends excluding senior management. “Execution happens at this level,” says Mulhall. “Including a director just stifles debate – the project team just The project team 4 It’s now time to develop the functional spec for the system – and here it’s definitely not about bigger is better. Think ‘less is more’ and keep it tight, covering only what’s required to deliver the business benefits required, and identifying but also separating out what are ‘must haves’, ‘nice to haves’ and ‘might want in the future’. They’re all important for future-proofing (you Functional spec Manufacturing Computer Solutions • Shortlist 2006 • www.mcsolutions.co.uk manufacturing business strategy system selection don’t want to be paying £1,000 per day for customising or new code some way down the line), but they need to be put in context. As ERP developer McGuffie Brunton’s sales director Howard Joseph says: “Don’t bother describing your business to the nth degree, but detail some specifics you will need, like you always need to do your putaways that way.” Mulhall adds: “There’s always enough for potential vendors to have a go at: they almost always come back and query some elements anyway.” Best advice: get away from the make-weight, boxticking and detailed software definition approach – and go instead for conveying understanding and context. The ITT remains important, ultimately forming part of a legal contract, but making it more litigationready than communicative helps no-one. Despite consolidation in the IT industry, choosing your shortlist is still not trivial: several of the ERP vendors now offer multiple suites with overlapping functionality, scale and industry coverage on different platforms. If you’ve employed consultants that are in the know, this is part of their role. They won’t send your ITT to everyone – only those they know have systems that are a likely fit and that have the right industry heritage. If you’re not using consultants (actually, even if you are using them) search our new website (www.mcsolutions.co.uk page 16) for information on ERP vendors’ systems and capabilities. You’ll find plenty to guide you, and you can search by functions, company names, whatever you need to know about). You can also buy the Benchmark Research report on user satisfaction scores covering a range of ERP providers (see page 14 or email enquiries@benchmark-research.co.uk). You might also talk to some of your suppliers and maybe some customers if the relationship works – see what they’ve got, what they think of it and why they chose it. And there’s the Best of British Manufacturing IT conference coming up on September 26 at Haydock Park: check out the website at www.bestofbritishmfg.co.uk and send some of your project team to find out what to look for and who might suit you. The bottom line is you’re looking to send your ITT to just six potential candidates to keep your selection process manageable and in a sensible timeframe. 5 Shortlisting suppliers out. Allman Horrocks recommends first inviting the shortlisted IT vendors into your plant for a preview day so they can talk to key individuals and get a closer feel for the business. Then you need a series of two-day demos where the spec becomes the script. Says Mulhall: “Insist that every point of the functionality is seen and scored in a structured fashion by the project team, and ensure users also complete scorecards.” It’s also time for visits to vendor reference sites, matched as closely as possible to the size, style and scope of your own company – and with your project team using the opportunity to go through similar detail where possible. Consolidating the scores – which must include rating IT suppliers’ people and culture (it’s subjective, but it is important that everyone feels comfortable with the people and their approaches) – determines the ranking. 7 Initial selection However you approach it, shortlisting is about looking for fairly detailed responses against the spec, also establishing whether what’s being offered is from the core package or add-ons that will need to be built or integrated, paid for and maintained. You’ll also want to see ballpark costings for licences, consultancy costs and so on to get an initial idea about how they’re shaping up. Use this information to pare down your six to just three by scoring the offers logically against your specified items. Then you’re into the demo phase for the final shoot- 6 But it’s not over yet: best practice advises the project team and/or senior management (depending on the company style and individuals involved) spending another day or so with each of the finalists at their ranch. In part it’s about clarifying final points of the functionality, but use the opportunity to check out the help desk, for example. There’s also a tacit assumption here that the vendors you’re looking at are financially viable and their systems likely to stand the test of time. “If all that confirms your one, two, three ranking, you now want to get to more realistic monetary proposals,” suggests Mulhall. “They understand your business now so there’s nothing to stop them putting together a firm proposal with the right mix of modules, numbers of licences and so on. Cost comes into play but it’s not always the cheapest that wins. We’re often asked to manage this side: we’re old and grizzled in this respect and we’re good at sharpening the pencil and boxing around with day rates and the rest.” 11 Final selection Manufacturing Computer Solutions • Shortlist 2006 • www.mcsolutions.co.uk How to get a good implementation Management basics The good news is you’ve already done a lot of the hard work during the system selection process. In laying the foundations for your new business and specifying and choosing the ERP provider you got your people working in a different, project-orientated, cross-functional, collaborative way. And that’s invaluable because these days, the reason for failure is hardly ever the hardware or software – it’s the same people and project issues you faced on the way to getting that system decision. That said, a good project does have common and identifiable ingredients beyond those you’ve already discovered about communication, involvement, the right people and adequate management sponsorship. They include: investing time and money in a proper conference room pilot facility for the departmental user teams; taking data cleaning very seriously; all the usual stuff about bite size project chunks and milestones; ensuring regular project review; and making problem resolution fast and efficient. As Richard Thomas, marketing director for K3, advises: “Have another kick-off meeting for the project imple- 1 key people in the business that understand their areas end-to-end and the objectives of the new system. Easily said and obvious perhaps, but those people need to focus on change as well as the detail. “Too often projects become technology-focused and change management happens too late. Change needs to be part of the project implementation with a change stream so everyone understands where the project design is going and its implications,” says James. And that needs to touch all stakeholders so that hearts and minds can be changed if need be, but also so that new roles, processes and jobs are understood and resistance is minimised while co-operation and good ideas are maximised. 3 “Expect the unexpected; plan for things that aren’t on the plan” Sarah Cobb, Moss Plastic Parts mentation team… Give it a name so it’s got a life of it’s own… Make it visible… There must be top level commitment for all the meetings… Choose a project manager that you know will make it happen.” But also: set realistic timescales, again taking account of your employees’ day jobs; set expectations and reinforce credibility by shouting about successes; have that risk assessment and mitigation plan up your sleeve; and keep project ownership within the business departments themselves. Doing all this well is less about sequences and more about processes – so what follows are important observations from those in the know. Yes the business analysis was done for the selection but you need to start again. The point is this is no longer theoretical: you’re about to put in a new system and change established systems and processes as well as real people’s roles and functions – and you need to make this smooth and re-usable, maybe across sites and divisions. Sarah Cobb, group IT director at Moss Plastic Parts, who has had considerable experience of systems and extensions, says it’s best to go back to the business objectives and use a standardised approach to build the project basics and the detail so that nothing is missed. “One of my overriding considerations with any new system is can it work in our big manufacturing sites and small global distribution sites alike,” she says. “Does that matter? What are we trying to achieve? The shape of the project also has to be driven by who is going to be involved. Do we need to go for a dictatorial approach to process delivery? Or will users be allowed to be innovative? Or are you going to computerise existing processes as a start point?” Hence the start again advice. It’s not difficult to design your new system on paper – or even on-screen. But how good that design turns out to be is revealed in the testing. You can do a lot to improve the outcome by getting away from assumptions or incomplete information about how things are done right from the start. “Never make assumptions,” warns Cobb. “Be aware of the capacity for ambiguity in the English language, and design your system to deal with things that aren’t meant to happen.” She agrees that it appears to fly in the face of proj- Start again It’s not an IT project This is, or should be, a business project so you still need total management commitment. Melvin James, enterprise service director at consulting firm Diagonal, describes that as meaning the board giving over part of its meetings to steering the implementation while the project delivery team comprises all the 12 2 4 Assume nothing Manufacturing Computer Solutions • Shortlist 2006 • www.mcsolutions.co.uk manufacturing business strategy system implementation ect management, but adds: “Expect the unexpected; you need to plan for the things that aren’t on the plan and get them on the plan, taking real life into consideration! Think, ‘If something can go wrong a user will do it,’ whether that’s just pressing a button to see what happens or sheer lack of understanding, or the fact that they forgot to mention the rare occasions when they do ‘x’ instead of ‘y’.” ect phases. James makes the point: “People aren’t used to changing requirements and weekly, daily, or monthly cycles. They’ll need a real level of management to help them though that.” One of the most common reasons for projects judged as failures is that companies underestimate and under-budget for educating and training their users. It may take IT and the project team several months to put the system in, so use that time wisely, with education and training in your conference room pilot so that when the system goes live, users are comfortable with it and understand what they are supposed to be doing, how and why. Cobb says this should also be ongoing long after the implementation is complete. “To get the best out of any big system you need regular education and training. A good way of doing this is to introduce some new stuff but also go back over basics. Look at help desk logs to see what keeps coming up to drive your subject coverage. Also sessions where users tell others shortcuts and tips they’ve learnt can be useful for spreading best practice. If you see something good happening in one department or on one site, then make sure you roll it out. But also be ruthless: if you hear something happening that you don’t want then stamp it out.” You need to consider your data up-front. Where does it come from? Where might there be holes? Are fields and structures being used differently across the business and its various sites and regions? How are you going to populate the new system? Cobb advises: “No matter how good you think your data is it’s worth reviewing it. Different software brings different aspects to the fore that may well reveal problems. Different sites will use different fields, like industry classification for example, in different ways and you have to bring that together.” Diagonal’s James adds: “There will be a lot of work: companies will find data missing and wrong, and doing this properly takes as much time as change management. Companies should allow 15% of the overall project effort and certainly not leave it to the end. People are always surprised by how much is involved and many projects that could go live technically are delayed by data that’s just not ready.” 5 Training, time, resource It’s not always applicable, but there is an argument for computerising at least some of your existing business processes before doing the full reengineering job. “If you put a basic system in for most of the business you can get it in quickly and risk-free, and maybe go for some easy wins first,” comments Cobb. “Then you’re ready to deliver advances and changes in the next phase.” She gives the example of preserving sales’ often complex pricing structures, or handling acquisitions where taking a vanilla approach to conversion and integration takes a lot of the initial risk and time out of the equation. A caveat she observes: “If you say you’re going to deliver advances in Phase Two, you must do it when you said you would – otherwise people you value will stop coming up with good ideas for the business. And to make that work you need a semi-formalised way of capturing ideas, reviewing them and a cycle of feedback.” 8 Computerise status quo 6 Get the data right Customisation Be very clear about where you spend your money, particularly when it comes to customisation. Says Diagonal’s James: “You need to understand the incremental cost of going beyond a baseline system. There may well be value in having some differentiators for your business but it will take investment in money and time, and potentially risk in the future in terms of keeping it up to date. It may also be better to leave this effort to a second phase.” 9 10 7 You need a robust project management methodology and a clear and very visible roadmap of how the project team is going to deliver over the agreed timeframe. More than that you need to keep communicating with people on the critical path so that they know and sign up to when and what to do in the proj- Project management lead the business project but they will need to be intimately involved throughout. These are the people that understand how the current system is put together, and these days more organisations’ IT folk have that blend of business, operations and IT knowledge that makes them especially useful. James: “They know how the data flows today and, since the business analyst skills tend to live in IT, they have the best knowledge for practical delivery of the new system to meet the business objectives. They also often have different but equally valid views.” And there’s another thought to ponder: you ought to think beyond the implementation itself to IT support – so, for example, provisioning of hardware and services. Do you want to build or acquire new IT skills if they’re needed, or does it make sense to move that over to a third party. “Be very clear about what’s being offered and the business case,” advises James. “Is there a genuine ROI, or is there a risk? Think about everything, including refresh rates of your technology.” I Enter 210 at www.mcsolutions.co.uk/enquiry IT department IT may not Manufacturing Computer Solutions • Shortlist 2006 • www.mcsolutions.co.uk 13