Stop buying “best of breed” technology…
IBM recently announced that its cloud platform team won a major deal to replatform several American Airline customer facing and back office applications into their public cloud. Apparently they beat out Microsoft and AWS in a cloud back-off this fall to win the deal. The announcement came during the AWS re:Invent conference, the timing wasn’t missed by the tech media either…
What Everyone Can Learn from American Airlines
In AA’s annual report they made several mentions of the difficulty of integrating applications.
While we have to date successfully integrated several of our systems…we still have to complete several additional important system integration projects. The integration of these systems in a number of prior airline mergers has taken longer, been more disruptive and cost more than originally forecast.
The implementation process to integrate these various systems will involve a number of risks that could adversely impact our business, results of operations and financial condition. New systems will replace multiple legacy systems and the related implementation will be a complex and time-consuming project involving substantial expenditures for implementation consultants, system hardware, software and implementation activities, as well as the transformation of business and financial processes.
Granted, airline operations are highly complex and integration failures can quite literally cripple a transportation system, the systems in question here are customer facing and back-office not operational.
The point is that integration sucks.
80% is Good Enough
Most companies think they’re business problems are unique. That belief causes them to invest in “best of breed” point solutions that address those pains in a comprehensive manner. The problem with that strategy is that problems don’t live in a vacuum. Solving one problem with a technology solution but failing to understand the relationship it has with the rest of your ecosystem is a recipe for failure. Implementations and integrations can last years and cost millions. By the time the project is complete that “best of breed” solution has lost its leadership position to some new technology or worse, the industry has changed and made it irrelevant.
So what if that vendor doesn’t have the “best” solution for your problem? My experience tells me that if a company can meet 80% of your requirements but also support related areas of your business at that same level then you are better off. PaaS is more than capable of picking up the other 20% of your use cases.
Take retail as an example. You could buy “best of breed” e-commerce, OMS, and marketing solutions from three different vendors.There is no single vendor that is the industry leader in each category. You would then need to integrate each one to the other at great cost and effort. The other option would be to buy from one vendor who has already done the integration heavy lifting. Sure, you only going to get one “best of breed” solution but the money and time saved is significantly more than any one system is going to earn in the short term. Trust me, they value proposition between #2 and #3 isn’t that significant in most cases.
Integrations of the Future
I’ve written about how SaaS is leveling the playing field today but many companies still miss the point. Don’t buy 10 different SaaS platforms without a defined and programmatic integration. If you are investing in a cloud solution, look for a vendor that has complementary solutions you can easily move into over time without investing millions of dollars or hundreds of hours setting up. Push-button integration is the path of the future.
I fully expect to see more platform companies in the future. Those platforms probably look more like NetSuite than Salesforce but that’s for another day.