As much as I would love the innovation happening in consumer tech to be rivaled in business tech, there are some good reasons why business has lagged behind.
Hierarchies of responsibility and shared responsibility. Someone is responsible for the decisions someone else makes. This may be a manager-managed relationship, or it may be a few steps removed, or it may be across departments, or with a regulator. Consumer tech is for individuals. If your calendar fails to remind you to take out the garbage, what do your neighbors care? Business tech is for teams, departments and the entire org. This is why “governance” is always an issue. Managers need to know what has been communicated to their people and what action was taken (or not taken) as a result.
Accountability. Will I be fired if I rely on an algorithm to alert me to problems? Will I be told that problems are still my responsibility whether I’m proactively alerted or not? Will I be scapegoated? The cost of a missed Facebook update may be a twinge of disappointment, or never ever known. The cost of a missed corporate social-network update may be a missed meeting, an angry customer, or losing your job. Governance again plays a role. If a system knows me as a person, then any activity related to me needs to be reviewable. And with so many systems distributed through the enterprise and the cloud, there is obviously a need to centralize this activity and make it searchable. And my responses and actions should be captured as part of governance by my management.
For these reasons I prefer deterministic solutions, and business data that captures the entire context. Intelligent, urgent alerts are good, but systems must maintain the complete context. Systems that urgently alert us must track everything they evaluate and show what did and did not cross the threshold.
Can a company go in the opposite direction? Can it “give up control” to probabilistic approaches? This company would understand that, in fact, it did not give up control and that probabilities are driven by feedback on their effectiveness. This hypothetical company would be a large optimization problem driven by measurement and feedback. But what would happen to lines of accountability? I don’t think the SEC or Sarbanes-Oxley would take a favorable view of probabilistic accounting. And the company’s interfaces with outside customers and suppliers will not be on equal footing. Data-driven aspirations will meet the friction of human activity, of the legal system, of “everyone else” who isn’t participating.
I don’t see these issues of responsibility, accountability and governance in business going away anytime soon. Business operations with clear interfaces can be hived off and run this way, as is the case with routine manufacturing. Small and mid-sized, privately-held companies are in the best position to take advantage of consumerized tech in business. This may be the most dynamic period in their history.