.svg){kind=small}
The amount of data an organization has is a Vanity Metric. Data moats are a reasonable argument, but a moat is a passive and defensive tool. The real contributor to an organization’s success is its ability to efficiently produce insights from data. An insight engine is a proactive and offensive tool that truly contributes to an organization’s competitive edge.
A modern competitive AI-first organization must have a holistic strategy and operating model that encourages the creation and improvement of insights. Similar to how product companies have systems that promote the creation and improvement of goods.
So what is needed to power the insight engine? What could this mysterious AI strategy look like?
The traditional operating model is divided along product lines, brands, and/or business units. This is a good model for a product-centric business. It organizes all resources in the same group for maximum output. The problem arises when the same organization wants to integrate insight-driven features into products. In this paradigm, the solution is to assign a data team to each group. Without proactive AI leadership, an organization can fall into a siloed structure.
This is a short-term solution that lets an organization claim they are an “AI Company” but this is not a scalable model for unlocking AI innovation. The data is fragmented, data governance is inconsistent, and data teams are cut-off. The very resources needed to power an insight engine are fractured.
The value of data isn’t linear. If you have twice as much data you have more than twice as much value. This is because the value is in the insights drawn from the relationships between data, not the data itself.
This realization has two very important implication:
Point 2 is the basis for calling data volume a vanity metric. An organization with a coordinated data strategy will outcompete, outmaneuver, and outperform its competition. Every product and service that such an organization creates will have the entirety of their data assets backing it; a powerful position to be in.
(Sorry, Spock)
Cybersecurity, data governance, and privacy are hot issues for any AI company. Fragmented datasets lead to fragmented data policies and a hopeless fight to solve those issues. Centralizing DataOps brings order to the chaos. A focused and consolidated DataOps strategy allows a company to effectively address concerns from customers and most importantly society.
Finally, these organizational data services are exposed to agile teams, products, and business units through a collection of standardized APIs and schemas. The game-changing advantage is decoupling the rate at which product teams need to grow and the rate at which data teams need to grow. For an insight focussed company, this is vitally important and provides a huge competitive advantage.
Being “AI First” isn’t about having the latest AI/ML features. It’s about having a holistic AI strategy that encompasses the whole organization, its people, and how everything executes.
Startups have an inherent advantage and can leverage their lack of legacy structures and organizational maturity to build a modern AI-first operating model. Larger enterprises need exponentially more vision and will to achieve the same results, but also have the resources. Microsoft’s Core Engineering, Google’s AI-First Initiative, and the famous Amazon Platform Mandate are three successful examples.
It's time to move on from data moats and data volumes. An organization simply needs technical leadership, a trusting team, and the vision, to successfully build out its insight engine. This is what will separate organizations that simply use AI and those that are AI-first.
With more than 25 years of experience in the travel industry, Mr. Waller has served as an advisor to Fetchy Fox, with a focus on strategy and deepening the team’s understanding of the challenges faced by concessionaires and airport operators.
.png)
