I've got that ERP feeling again. However, I refuse to start another article on the subject with words such as "disappointment," "failure" and "unfulfilled expectations." (Whoops, I just did!) No, this time I'm looking for the positive (it must be all those Gartner and AMR figures predicting IT spend growth this year going to my head.) 

Of course, in the midst of my research I still have to wade through pieces whose first paragraphs re-tell horror stories of chronic and sometimes terminal ERP installation failures and I smile wryly at a sample line of defence from vendors, namely "you've been doing it wrong." But then something else strikes me - a pleasing consistency in the advice and implementation methodology now offered by those involved in the ERP market place. A clear sense of lessons learned and expectations firmly under control.  

Here's a selection of the best and most practical advice available.

Getting started

Andy Makeham, CEO K3 Business Technology Group, member of the ERP Vendors Forum (www.erp-vendors.org.uk) and author of "How to guarantee successful ERP implementations," cuts straight to the chase in terms of why ERP implementations have a history of failure: "The problem, as ever, lies with people." He simply points out that if ERP is about sharing integrated data, operating as a team and working towards common business goals, then implemeting an integrated system in a non-integrated company is doomed  to failure: "Interpersonal relations are the rocks on which smooth information flow founders!" 

So unsurprisingly it's a people thing and while we're talking about people, Phil Robinson founder of BPIC (Business Performance Improvement Consultancy www.bpic.co.uk) places the emphasis for successful implementation firmly in the hands of management: "good software helps but is not a substitute for good management."

Both are agreed that good ERP implementation begins right at the start of the selection process. By involving a team of people at the outset, understanding the business issues and the scope of the project in terms of the potential benefits as well as the resources that will need to be committed, those people will feel part of the decision process. The key to any successful implementation is the "transfer of ownership" from the system vendor to your staff, and you can only achieve that if your team feels involved in the process, and the earlier the better. 

Doug Miles, Marketing Manager at mid-market ERP supplier Infor:swan (also members of the ERP Vendors Forum) endorces the importance of those early stages: "Usually people start detailed comparison of ERP systems too soon. What they need to do first is decide which key areas or processes in their business are most in need of improvement, and set themselves some improvement objectives they can measure."

Teamwork

So the roots of successful ERP implementation lie in teamwork and measurable objectives, but who makes the team?

An overall project sponsor is ranked by Andy Makeham as "a critical success factor."  His view is that companies need to have commitment right from the top. In most SME manufacturers this means visible and positive involvement from the MD in meetings and with the steering committee.

A project manager with "company wide credibility, determination and a good knowledge of the business" are the characteristics to look for according to Phil Robinson, who also stresses that the project manager's access to the project sponsor when necessary must also be assured. Andy Makeham goes as far as to say that the selection of this individual "should hurt" because it may temporarily take he or she away from critical day-to-day duties.

Then there's a steering committee, comprising key managers or department heads whose main role is to pull together the overall objective or vision for the project and subsequently the more detailed "new ways of working" that most successful ERP implementations now seem to require - more on this later. 

The steering committee work with process teams, representing all functional areas who are required to examine existing processes and help establish,document and measure these new ways of working.

Andy Makeham for the ERP Vendors Forum and BPIC's Phil Robinson both refer to a "Conference Room Pilot" as the cornerstone of the ERP implementation. This takes the form of a room containing some system terminals to simulate all the different company functions from forecasting and sales order processing through processing to shipment and financial reporting. Initially the pilot is used by the project team to try out their ideas, but as many members of staff as possible should be encouraged to get familiar with the system so that they are comfortable with and have documented how they will carry out all their tasks before going live.

Communication

All this intense strategic activity will be a little unsettling for employees without correct communication of the overall project. Uncertainty breeds fear, particularly when a bit of new technology is thrown into the mix. Andy Makeham urges communication to play a large part in any ERP project, and to all levels of employee. He suggests giving the ERP project a name and issuing information about progress in a newsletter, on a notice board or on the company intranet.

As well as progress, communication also needs to address the "what's in it for me factor," by connecting the ERP implementation to improving the achievement of business objectives. Similarly, everyone else must see how their individual objectives will be easier to achieve if they take the trouble to understand and contribute to a successful implementation.

Ultimately, whatever good communication can't deliver in terms of cultural change, new performance measures that support a new way of working can.  (Abandoning old performance measures that support old bad habits is equally effective, yet more tricky to implement.)

Clear objectives

Andy Makeham observes that a lot of ERP implementations "never finish... they just drag on and on." He explains that this is because "at the start of the project no one defines the end conditions. How can we know when a project is finished if we don't know the end conditions! " What's called for is the setting of clear business rather than technological objectives, that are preferably quantitative rather than qualitative.

Rob Murphy of strategic consulting group, Atos Origin (www.atosorigin.com) picks up on the all important measurement angle: "the benefits are agreed and understood at project start, their delivery tracked through the project life cycle and measured after implementation to prove they have been delivered."

It's not inconceivable that third party help may be required when aligning system requirements with business objectives and here's where choosing the right sort of help becomes critical to ERP implementation success.

Choosing a partner

Many companies try to implement systems without the help of people who have proven experience in implementing such systems. It is possible to get such help without losing ownership of the project and from a third party who can be a coach without necessarily being a player. But Infor:swan's Doug Miles does have some words of warning: "If you use an external consultant, they should be presenting the ERP vendors with their overview of your business requirements, not translating that into a 300 page ITT tick-list."

His are not the only words of warning. A recent piece of research from analyst firm Ovum highlighted a huge variation in the cost per seat of ERP systems depending on how the software was implemented. With the reported range of cost being between £2,000 to as much as £58,000 and with implementation on average accounting for 90% of overall costs, future adopters might be keen to discover what was found to be the cheapest way to implement ERP.

Well the golden rule seems to be to avoid software customization: "stick to vanilla business processes" Ovum advise. The research also found significant price differences between tier one suppliers such as Oracle, SAP, JD Edwards and Peoplesoft and tier two suppliers such as Exchequer, Sage and Microsoft Business Solutions and that it was more cost effective to use consultancy services provided by the software maker than use one of the big four IT consultancies. I'm sure none of this comes as a huge surprise but either way, Ovum warn that licence fee alone is a poor indicator of implementation costs. 

Even though employing one of the big consultants is seemingly beyond some companies requirements not to mention wallets, adapting their best practices is not out of the question.

In November 2003, Cap Gemini Ernst & Young launched EDPI or "Event Driven Package Implementation," an approach that: "incorporates innovative techniques to ensure that all project stakeholders fully understand the way in which their ERP project is capable of helping companies achieve their business goals." Based upon the "3Cs" - "Criticality, Collaboration and Culture," already you can see that they have little to do with specialist IT training and everything to do with understanding business strategy, implementation partnerships and unique client cultures and conditions. The same issues raised by organization's like the ERP Vendors Forum.

According to Cap Gemini Ernst & Young: "organizations often fail to secure the anticipated benefits of an ERP project due to failure to align the project to business goals and therefore failure to achieve full stakeholder buy-in" (the "transfer of ownership" that K3's Andy Makeham refers to.) "Risks and issues that are often swept under the carpet causing future disruption and delay in delivery" are teased out through a series of face-to-face meetings (Accelerated Solution Environments, Collaborative Consulting and Rapid Solution Workshops to name three) that "massively accelerate and enhance the pursuit of common goals and resolution of issues and problems that typically cause ERP project delays and stakeholder alienation." Sounds similar to a conference room pilot. Wordy they may be, but not inconsistent with other views expressed.

Choosing a "flavour"

Partnerships aside, there is also emerging consistency over the question of adopting a "vanilla" business process approach to gain maximum benefit and rapid installation. In other words don't bend the ERP system to suit your old way of working. Instead take those best-of-breed processes and adopt them throughout your organization. Use those processes to maximize the efficiency of your company and reduce the need for customization. This might seem like a backward step, but sometimes you have to take a step back to move three steps forward.

It's an interesting dilemma. There used to be and probably still are those that say software systems should not make you change the way you do business as an organization, that functioning groups are not easily or quickly persuaded into radical change, particularly if imposed from the outside. It seems to me that implementing ERP without process review potentially means maintaining bad process but now it becomes automated bad practice - different tune, same piano.

In fact, perhaps if early ERP adopters had measured the benefits achieved from changing the way the company did things (ie. identifying duplication and waste)  alongside those gained simply from automating data processing, they would owe ERP a greater debt of gratitude for exposing the shortcomings of underlying business processes, masked by former boom years of increasing revenues and profits.

So, maybe adoption of best practice needs to be accepted and the only processes worth protecting with your corporate life are those unique processes that keep you ahead of the competition. Other commodity processes or those that simply seem unique but are only unique because of the nature of the business sector in which they operate are ripe for examination and improvement.  Perhaps the business requirements are only one aspect here. The availability of management and technical skill is equally important and may be the most critical factor in deciding which route to take.

Where are we at?

Christian Verstraete, VP of HP's Manufacturing Industries Practice, is clear that for many companies, having basic ERP is no longer a competitive advantage, but not having it is a distinct disadvantage. On basic implementation his advice is to keep it simple, adapt processes rather than the package (or face high costs, migration issues, low benefits etc.) and choose your implementation partner according to your needs. But whatever you do, don't let an implementation take forever because while its going on, so is life. Take a rapid, focussed approach and move on to the newer challenges of ERP: add-ons such as planning systems, supplier relationship management and CRM/sales force automation.

But remember an ERP system is only the tool you need to achieve that step change in business performance, it is not (despite what it says on the box) a solution in its own right.