Tier One ERP solutions can be 4,000% more expensive than Tier Two

International corporations ought to adopt a two-tiered enterprise software (ERP) approach for their business and manufacturing IT strategy if they want to keep costs down and get implementations in months or weeks rather than years. That’s top line advice from independent analyst Dennis Keeling. Research commissioned from him, by ERP software vendor Scala last year, reveals average ERP prices at eur 15—30,000 per seat (full implementation cost) for Tier One ERP systems, compared with eur 2—8,000 per seat for Tier Two systems. It’s a huge gap, and Keeling says subsequent research, also by him, shows that in some cases it’s far worse. “My latest figures now show Tier One total implementation costs at eur 5,000 to eur 87,000, while Tier Twos are eur 2,000 to eur 10,000. And those Tier One numbers are for implementations essentially in Europe; they hadn’t even attempted to go to more difficult countries.” His latter figures come out of an in-depth study of 40 very large international enterprises, typically with manufacturing in multiple plants in several countries, an HQ in another and distribution, sales and support in several more. The earlier results are from research with 10 such organisations. “On the run up to Y2k, most of these companies initially attempted to go wall-to-wall Tier One ERP,” he says. “But all of them weren’t achieving the time scales, and found they had to go for additional Tier Two ERP vendors, at least for some countries, and sometimes more than one, to get over the Y2k wall.” Reasons included: cost of implementing a Tier One solution due to complexity; cost of deployment and support; and availability of infrastructure in Third World countries. Other problems cited were: stand-alone solutions are more suitable than centralised for local SME operations; localised solutions are easier to implement; and standardising on two vendors helps application integration. “The survey shows that organisations that have adopted a two-tiered strategy have successfully managed to implement their Tier Two products quickly, even in the most remote countries. At the same time they still achieved the control and standardisation they originally aimed for with a single tier solution,” says Keeling. Many have long suspected substantial differences in total cost of ownership (TCO) across the size range of ERP solution, but there are problems in attempting comparisons. For a start, getting big corporations to really open up on real costs is a tough nut to crack. Beyond that, there are complexity and scope issues that make comparison questionable. However, the numbers are easily big enough to be statistically significant. And Keeling argues strongly hat he achieved like-for-like validity, claiming his methods were rigorous throughout and emphasising his presentation of TCO results for implementations only in comparable countries and circumstances. Fact is, the sheer scale of Keeling’s documented and detailed divergence cannot be ignored. Keeling suggests that at Tier One, ERP software licence fees can be as little as 10% of the ultimate TCO, whereas at Tier Two, it’s in the 50—70% region. “Tier Ones have had to be all things to all people, so they’re sophisticated and complex, requiring lots of configuration,” he says. “Also, they use third parties for that and they’ve got a rather different vested interest. But tier Twos mostly do the implementation themselves and they’re all about getting in and out as fast as possible.” It’s a good result for Scala and for the rest of the mid to small range ERP vendors. David Topping, senior vice president of marketing at Scala, says: “With the negative awareness of failed projects, expensive implementations and customer dissatisfaction, we were keen to prove that a new strategic approach to ERP could successfully deliver on the ERP dream of integrated IT systems.” One has to doubt, however, whether research of this nature would have been sponsored by his company had there been any doubt about the outcome. Nevertheless, Keeling identifies some useful, if unsurprising specifics. He says that while Tier One systems are ideally suited to cope with the complex needs of centralised functions and huge numbers of users, they are not most appropriate for the less complex needs or localisation requirements of a branch or sales office in remote countries. This, combined with the total cost of implementation and on-going support of Tier One systems, versus the ability to keep local costs firmly in control by using local implementation partners with a Tier Two system, provides an extremely strong case for a two-tiered strategy, he insists. The key to driving success and lowest costs out of that is going for standardisation in a two-tier system roll-out, with a policy of using a company-wide implementation template coupled with built-in localisations rather than ad-hoc customisations. And on the face of it that limits you somewhat. Keeling notes that at Tier Two there’s not been the rationalisation that we’ve seen in Tier One to allow standardisation. He names only Scala and Systems Union as adequately multi-lingual, multi-currency and localised to allow a genuine standard two-tier global implementation strategy. Scala, for example, has localisations for 35 countries and moves up to 36 languages in the next May release. Topping claims the firm has implemented in 140 different countries and that one project alone involved installation in 75. Other ERP vendors, says Keeling, don’t have the financials and regulatory localisation footprint, or are rolled out through too many distributors with local customisations to get the full two-tier cost- and time-savings. But he concedes that there is also a powerful argument to consider the spread of similarly able manufacturing-specific ERP systems at this level, for example, in terms of their regional, manufacturing style and global industry-specific coverage.