Skip to content

Chapter 2: Why Build An Enterprise Knowledge Graph?

  • EKG Use Case Taxonomy
  • Cost-Benefit Analysis
  • Metcalf's Law and Network Effects
  • Business Value
  • Measuring EKG costs
  • Measuring EKG tangible benefits
  • Measuring EKG intangible benefits
  • Brains are prediction machines

When we build any large resource for an organization we are going to need to ask for money to build it. If you work for a company like Apple and you are respected expert at building enterprise knowledge graphs, you many not need to go to a finance committee and justify your spending requests. Apple hires world experts in specific fields and then trusts their judgement. As long as you have a good track record they will keep giving you funding without asking a lot of questions.

But most of us don't work at Apple and we are not world leaders with a long successful track record of building EKGs. We will have to fight for every penny we spend on our EKG pilot projects until the ROI is so clear to everyone that the finance people are begging you to expand the scope of the EKG.

Until that happens, we are going to need to learn to speak the language of finance to get our EKGs off the ground.

EKG Use Case Taxonomy

A Taxonomy of Graph Use Cases

  • Performance Fast Relationship Traversals (no JOINS) Easy to write complex queries ** Easy to perform deep link analysis
  • Flexibility
  • Rules

Metcalf's Law and Network Effects

Cost-Benefit Analysis

[The Business Value of Computers] Wikipedia on Business Value Peter Drucker

Measuring Costs

  • Cost of extracting knowledge from operational source systems
  • Change data-capture (CDC)
  • Publishing business events

The Easy to Measure Benefits

  • Integrated views of anything
  • Integrated views of your customers
  • Centralized business Rules
  • Recommendation systems

Difficult to Measure Benefits

  • The value of insights
  • Tracking early insights
  • Predicting future insights
  • How are innovations and patents valued in your organization?
  • Focus on datasets that generate shared value

Brains are Prediction Machines

Pick up a ball. Throw it in the air and catch it. A simple act. But to perform this simple act our brain has to do incredible things. Most importantly, your brain will predict where the ball is going so you can move your hands to catch it. And it does this task effortlessly in real time. You don't really have to think about it.

This metaphor, of the brain as a prediction machine that easily can predict the future position of a ball in flight, is how we want you to think of your enterprise knowledge graph. The EKG is a machine that will help you predict what actions you can take to best serve your customers. That is one of the best reasons to build them.

We use the metaphor of the EKG as your companies "brain" carefully. Because the current generation of graph databases really are very different than a human brain. Neurons in a human brain are very different from a CPU thread stepping through vertexes and edges of a graph. A typical enterprise graph has about six connections between a vertex and neighboring vertices. A human brain has around 10,000 connections for every neuron. An EKG can be inspired by the human brain just like plans are inspired by birds and submarines are inspired by fish.

Systems Thinking Questions

  1. Should business units be charged for use of an EKG? How would change-backs be calculated?
  2. What is the role of innovation in your organization. Does your organization manage a list of ideas from all employees? Are the ideas sorted by cost-benefit analysis?
  3. How would you describe key events in technology and their relationship to cost-benefit analysis?
  4. What is the relationship between the Chief Executive Officer and the Chief Financial Officer in your organization? Are finance staff trained on valuing intangible benefits?
  5. How is overall knowledge-capture valued in your organization? Do employees get a bonus for sharing codifiable knowledge?

References

  1. Waste in the US Health Care System Estimated Costs and Potential for Savings - EKGs could potentially save hundreds of billions of dollars a year using evidence-based data to find inefficient healthcare procedures.