Choosing a major ERP platform is a huge decision, and one that firms don’t like to make too often. The stress of moving sensitive corporate data and varying software plug-ins from one environment to another - without a devastating loss of service - makes moving house look simple. Amsinck goes one step further, likening it to a marriage.
“When you choose a huge ERP vendor it’s like a marriage, whichever girl you marry you make a choice and you try to make the best of it. But if things turn bad you don’t invest more in it. Where will growth [for SAP] come from then?”
He explains that companies will go with a different approach once their frustrations with SAP boil over - perhaps allowing other applications or cloud services to connect to the SAP core.
“Then SAP will be out in terms of being an attractive vendor to businesses,” Amsinck concludes.
SAP’s licence fees are another area for concern. Amsinck describes a situation where he wants to experiment with a new CRM system, providing it to a business unit to try out for a couple of years.
“Imagine you take the whole thing as a pilot, buying licences, add-ons and other fees, but it ultimately doesn’t work out and you decide not to continue. Now the SAP pricing system says that’s fine, they take back the licences, but you’ll still pay maintenance even though you’re not using anything.
“You’ll pay 22 per cent per year - and to me that’s very weird. I’m giving the licences back, but they’re charging maintenance forever. So it’s cheaper to buy the same box from third-party vendors because they can’t charge me for the maintenance if we quit.
“So now, if all else is equal, if I can find a shadow IT product offering the same service, I’d buy that instead because the licensing terms will be far more attractive.”
He complains that this inflexibility from SAP makes it impossible for him to deliver on his CEO’s ambition.
“How can I live up to the CEO’s request to be more adaptable if anything I experiment with will haunt me for-ever? Am I the only CIO to worry about these licensing terms?”
Amsinck is now reluctant to add further SAP services to the core, saying that who supplies the peripheral systems that plug in to this core is now a very open question.
“For the first time in my career that’s where cloud comes in. Plug it into the core and you can get a very good user experience.”
However, Amsinck is not a wholehearted advocate of the cloud, due to the difficulty he perceives in integrating disparate cloud offerings into a consistent and joined-up system.
“We’ve looked into SaaS, but sometimes when you need heavy integration, it’s not the best option. We moved away from SuccessFactors and put our HR and management systems into SAP just when SAP bought SuccessFactors. I think the more integration you need, the less attractiveness there is in the cloud.
“Integration is the main reason you don’t find SAP in the cloud today. You don’t see anyone composing their enterprise platform in the cloud, you can’t integrate everything into one platform. When you have vast volumes of data in your pipes, and you’re connecting it all into one big enterprise platform, you need a clear and consistent data model underneath.”
The challenge Amsinck sees is that every software service comes with its own data model, and they’re not compatible with one another.
“We’re a long away from doing something serious with SaaS. It works if it’s really granulated, like checking emails for viruses or blocking malicious websites, but we wouldn’t use it for a key central service. We will look into it more as services mature, but it’s important that any service links directly to the core, as the master data will sit there.
“You can still load data in and out, but the value is in the service provided to the user [as opposed to core functionality], so you can extend your core platform with cloud services, that’s how to get most value with cloud.”
Despite what he sees as the cloud’s limitations, Amsinck is considering a new disaster recovery concept for LEGO that would be heavily supported by cloud services.
“We’re designing a new datacentre and backup datacentre, and I’m thinking how much can we just failover to the cloud?
“It would be a cost saving as there’s not enough space in our current set-up to do a full backup of the primary datacentre. If we can avoid buying more space or moving to a larger site, that will save money. I think the bigger question is if you can only pay for services when you use them, for example when primary systems fail. That’s when the cloud becomes attractive, but it’s less enticing when you have to pay a regular service fee.”
Sometimes, the power of the mainframe is the most cost effective answer. Computing's Peter Gothard puts Computing's readers' questions on the future of the mainframe to IBM's Z13 expert Steven Dickens.
This Dummies white paper will help you better understand business process management (BPM)