I joined Gartner in 1995. I remember clients would call and say they were getting ready for next-gen applications, but that they felt they had "enough time". The deadline in their case was, of course, the Y2K "wall". By the end of the year, the urgency grew and for the new few years, for several analysts like me, all hell broke loose.
Gartner had an elaborate call tracking and scheduling system called GAMEC to facilitate 30-to-60-minute consultations with clients. The average, annual GAMEC metric for an analyst was about 500 inquiries. In the 4 and a half years that I was at Gartner, through the end of 1999, my GAMEC count was over 7,500!
Having come from Price Waterhouse where I ran a software intelligence group, I handled many of the ERP software queries around SAP, Oracle, PeopleSoft, JD Edwards and other client/server vendors. Having done multiple implementations, I also handled many of the ERP systems integration queries around Deloitte, IBM and other SIs. And for years after leaving, in my own firms I helped many of these customers source and negotiate their ERP application management, hosting and other contracts.
As an industry, we declared victory that Y2K was not as lethal a disaster as some had projected. The reality is many clients waited too long and then paid significant premiums to SIs and still had poor implementations. Others did not do enough due diligence on their software vendors and many are still paying the price. Many only replaced back office systems, when updating operational areas would have delivered them much higher payback.
I get the feeling we are about to see a similar spike in interest in next-gen apps in a move that Hemingway called "gradually, then suddenly". As I tell my advisory clients "fool me twice, shame on me" - we made a lot of mistakes in the 90s and we would be remiss to repeat them.
I have had conversations with several SAP customers in writing SAP Nation 3.0 and others since it was published in March. Many of them say they are looking at migrating to S/4HANA, but then they add "we have time till 2025", the announced SAP end of life for their current ERP software. Honestly, they should be looking not just at S/4 but other new SAP technologies like C/4, IBP and Leonardo. BI customers should be looking at the Analytics cloud. BusinessOne customers at ByD and so on.
But, wait a minute, we don't have a hard Y2K wall this time, so what's the hurry? If anything, it is possible SAP may extend ECC support beyond 2025. Customers of Oracle, Infor and other on-prem vendors show even less urgency in looking at next-gen solutions. But the management changes and layoffs at SAP earlier this year reflect an industry where open source, ML, IoT are the future not ABAP or Basis. The acquisitions of Mulesoft and Tableau at Salesforce, the autonomous database and cloud infrastructure investments at Oracle are signs of a significantly changing industry. Your Board and management team are talking Digital Transformations. Do you really want to be stuck in 2025 with technology and talent from the 90s or even earlier? Have you considered your recruiting brand for younger IT professionals if you want them to support Baan or PeopleSoft? Have you considered the compliance and security risks with your customizations around that technology?
If we learned anything from the 90s, we should be preparing for things very differently
a) Spend time to envision your next-gen application portfolio
In the 90s, applications consolidated many departmental solutions and allowed companies to set up shared service centers across global subsidiaries. This time around the goal should be about (and the payback will come from) retiring last-gen customizations and tons of spreadsheets, consolidating many of the surround cloud solutions companies have acquired in the last decade, and most critically modernizing industry and geography specific operational systems. They were neglected in the Y2K rush and in the twenty years since every industry and country has gone through seismic changes. Evaluate your current major software vendors to see if they can support you (make them share detailed road maps for industry and geographical growth, and negotiate penalties if they delay planned deliveries), but also be prepared to look around. There are plenty of vertical and geographic application specialists, and you should consider custom developed applications which leverage next gen ML, multi-tenancy and other platform capabilities from Amazon, Microsoft, Google, Alibaba and others. All this will take months, if not years to evaluate.
b) Spend as much time to review your service providers and your own talent needs
No question - the last wave of services procurements was poorly done. Companies did not plan for a life cycle of needs. They hired SIs for initial implementations, then outsourced their infrastructure to hosting firms, then hired application management, often offshore firms to help with on-going care and feeding of their applications. Each of these was done over years, often with different providers. It has also become their biggest single cost, without proportionately positive results. However, so much has changed in the last twenty years. Services should be a lot more automated now. They should deliver shared service/multi-tenancy efficiencies. There are third party maintenance firms for level 3 support, rural sourcing and near sourcing firms for AMS. We are also seeing a reversal and several CIOs are insourcing talent. Much of it is for making their own products smarter, IoT compatible and contemporary in design. As with vendors, customers are having to look for talent with open source, ML, integration and other capabilities. Reviewing the changed talent and automation marketplace similarly deserves months of analysis.
c) Do more scenario planning, not traditional TCO/ROI analysis
I hear from several customers they cannot justify a move to a next-gen product. I can certainly empathize, but often their analysis of the cost and benefit is one-dimensional, focused on a single application and does not cover the lifecycle. We need a new generation of IT financial analysts who can perform scenario analysis to handle the complexity of multi-year, portfolio wide migrations.
d) Develop a new intelligence network
The one big difference since the 90s is proliferation in information sources. Blogs and vendor marketing materials are plentiful. But they tend to be anecdotal, or single threaded. And many tend to focus on the negative. Google the term "SAP Indirect Access" and you get 6.5 million hits. Is that a concern? Of course, but one of 00s of things to consider as you look at next-gen applications. Peer networks are a far better source of intelligence and sharing. Again, you need multiple conversations given the complexity of the portfolio.
Long and short, it's not too early to start reevaluating your application portfolio. Don't let vendor end-of-life projections lull you - or in reverse, force you to join a stampede in migration. Do it on your own terms and speed, but don't ignore the history of the 90s. You may be doomed to repeat those mistakes.