I recently attended Infor’s user conference: Inforum 2018 in Washington DC. Even though I couldn’t stay for the Lenny Kravitz concert on Wednesday night, I did pick up some cool ideas. Below are some of the most interesting ideas and some follow-up questions they inspire.

Delivering an Industry in a Box

Infor takes a very high-level, enterprise-wide view of the business world. They don’t want to just sell you a database or even a data warehouse. They want to deliver a fully functioning business and technology ecosystem that is fine-tuned to your industry. They call it delivering “An Industry in a Box”. They do it by pulling together all the pieces and delivering a complete solution with best-in-class data models, KPIs, and best practices for a variety of industries including healthcare, retail, manufacturing, and even fashion.

The strategy seems to be working. A Senior Director of Applications Development at a large retail company summed up the shift in their corporate strategy by saying roughly:

“We used to ask whether Infor was the best solution available. Now we ask whether Infor will meet the company’s needs. Unless there really isn’t a fit, we don’t look beyond Infor for a solution. I just can’t emphasize enough how hard it is to integrate all of the different systems and different vendors. It seems like it should be relatively easy but it isn’t.”

This is craziness, right? You’ll get locked into one vendor and be held hostage.

You’re probably right. Lock-in is still pretty much inevitable but business is all about efficiencies and there may now be factors like time-to-market that upend the wisdom of the past. Let’s consider some objections to this ‘industry in a box’ solution and how they might be overcome.

Won’t this Require Overwhelming Customization?

When companies have tried to provide a complete industry solution in the past, they have often still required a lot of customization. Their customers often felt that their core differential advantage needed to be expressed through customizations of the standard product. This attitude may be changing for a variety of reasons:

• These customizations sometimes resulted in more lines of code than there were in the original software product. Every time the product was upgraded it meant that the customizations also needed to be upgraded or even rewritten. This was a huge cost in terms of FTEs and time.
• Speed to market is now becoming more of a differential advantage than customization. Using best-of-breed offerings from a vendor who is dedicated to your industry means that they can incorporate best practices better than any single company could. And do it much faster. A customer’s unique customizations may be less valuable than getting the new software release into production in 40 days rather than 14 months.

Won’t this Result in Vendor “Lock-In”?

Yes. Absolutely. You’ll be betting on one company but you will always have the opportunity to replace them whole cloth or piece by piece when they are no longer at the top of their game. If, however, you have not customized your applications too much, it will be much easier to bring in another vendor in the future. The value of rapidly deploying a functioning and integrated solution based on industry best practices may overwhelm the fear and costs of being locked-in to a vendor. With everything on the cloud these days it is also getting easier to move from one vendor to another.

Going the “Last Mile” for Application Customization

One concept that Infor talked about that is critical to "industry in a box" was extending the effort to go the last mile.

In the video game industry, it is well understood that after you have done 90% of the coding and hard work for creating a new game that you are only achieving about 50% of the fun - and 50% of the revenue. That last 10% of effort consists of fine-tuning what was built, but it doubles the value.

Infor is committed to putting the extra effort into customizing their solutions for particular industries and particular applications. What they call ‘going the last mile’.

Focus on Training

Infor CEO, Charles Phillips noted that there are now more jobs than there are unemployed.  The Wall Street Journal recently reported that there are currently 6.7 million unfilled jobs but only 6.3 million unemployed. These statistics don’t fully capture those who are not officially considered to be ‘unemployed’ by the economists. There are also 39% of Americans who are employable but currently out of the workforce. Many are capable and willing to work but they are being kept from being employed by a lack of required skills. And that skills gap is only getting worse as job requirements become more complex as time goes by.

Gallup further reports that 65% of people who are employed are not fully engaged with their jobs. This lack of engagement results in a loss of productivity valued at $400B. When these workers were asked what would make them become engaged in their jobs their top answer was better and more training (increased salary was only #4 on list).

Both of these problems can be addressed with better training and overall education. Infor is taking this challenge head-on by building e-learning courses, and embedding gamification opportunities and employee encouragement features (they call it “Rave”) into their Human Capital Management system. To me, this is revolutionary. Everyone knows that the skills gap is a problem but Infor is uniquely focusing on solving it by delivering training embedded into fully integrated HCM solutions.

Is AI Better Embedded in Context?

Yes. Infor has an AI initiative called Coleman, which is ubiquitously distributed throughout the product. To date, it consists of easy-to-use user interfaces built on top of Amazon’s AWS services, such as Lex and SageMaker. Infor is happy to let Amazon (and eventually probably Google and Microsoft) build fault tolerant cloud infrastructure that they can then build on top of rather than recreate.

Infor is focused on delivering AI to business users who aren’t PhD data scientists. This strategy is similar to that used by Salesforce with their Einstein offering, which embeds predictions directly into applications. This allows them to deliver AI recommendations in context at the point where end users most need guidance. These AI recommendations are mostly delivered as suggestions that can be easily overridden or ignored by the user if they disagree or feel uncomfortable. This provides an easy, non-disruptive way for cautious business users to ramp up on AI without having to go all-in right out of the gate.

The Big Picture for AI => Get Worried

There is something profound occurring with Infor and with the world in general as we embrace AI. Infor delivers a product that runs much of companies’ day-to-day operations. This means they cannot only deliver an AI-driven prediction but can execute it on the same screen of an app. Their products can, for instance, not only predict an inventory shortfall a month into the future but also guide the user in requisitioning more product from that same screen.

This is a fundamental change in the use of AI. AI is most powerful when sensory input (big data) is tightly integrated with thinking (models and predictions) and execution. And this is where maybe, just maybe, we should think a bit about the machines taking over. In the not too distant past, AI systems were detached from the real world both in terms of gathering input (they couldn’t look at photos or sense temperatures in an engine in real time) and taking action. Now those barriers are coming down.

The next step is for AI systems to start their own experimentation. Specifically taking action, not just based on previously collected data, but specifically to learn about the world and generate better data. This will produce a virtuous upwards spiral (at least for the artificial intelligence).

When this happens we can just kick back and watch our businesses self-improve under the wise and benevolent guidance of our AIs. We just need to make sure we keep that off switch easily accessible.