Everyone is now a data + AI user.
The Four I’s of The Destination Intelligence Arc.
Initiate
Internalize
Integrate
Interoperate
In the age of AI, everything is data. And, data is money.
Most organizations think about data in terms of more is better. Bring a better dataset and a better tool, and you
are set.
The idea that more data and better tools lead to better decisions is starting to break down. Working with travel and tourism organizations, I see firsthand that the real bottleneck is translating data into context-rich insights. Organizations are swimming in data. Partner records, meeting recordings, chatbot conversation logs, mobility data feed, and customer reviews. When AI generates more data than one can comprehend, the real constraint has become turning all of that into:
Insights without needing a data scientist in the room
Something machines can read and process
Since data is money and everything is data, it leads us to the question:
“Can we absorb and interpret the data without calling on data experts every time?”
Traditionally, most travel and tourism organizations, especially at the DMO level, had a data, insights, and research team. Data lived behind closed doors. If you wanted insights, you reached out to the data team. The team had technical skills the rest of the organization lacked. Today, the gatekeeping function is changing. Access has evolved from raw data to refined intelligence.
Ideally, in an insight-driven organization, a marketing manager can ask nuanced questions about campaign performance and find answers without needing to understand data science principles. Data is not locked behind a gatekeeper. It breathes and circulates across teams and disciplines… Marketing, Sales, Advocacy, Community Engagement, Stewardship, and so on.
AI is a tool that can book, route, negotiate, and re-optimize on your behalf, but it requires a different kind of data infrastructure than a tool that drafts your emails.
The current wave of AI has accelerated this shift toward self-sufficient data users. These AI tools can act on our behalf, complete workflows, and make decisions for the organization. The difficult part is preparing teams for how AI changes how work gets done.
AI adoption is accelerating. Where organizations differ is how deliberately they use data, redesign processes, and deploy end-to-end workflows.
Destination organizations serve many stakeholders. Having worked at one for almost a decade, I saw that the DMO's mission is inherently distributed. Partner relationships, stewardship, branding, sales, advocacy, community engagement, product development, partnerships, often across geographies.
Each DMO function asks unique questions and generates its own data silos. Fragmentation is the default.
The Destination Intelligence Arc.
To navigate that complexity, we developed The Destination Intelligence Arc. A four-stage framework for understanding where your organization is on the data + AI journey. We distilled it from our work with destinations
and from a firsthand understanding of how AI is changing the way work gets done and data is used. We call the stages the 4 I’s:
Initiate: Doing existing work faster
Internalize: Shared knowledge surfaces
Integrate: AI handling end-to-end workflows
Interoperate: AI becomes an ecosystem
This four‑step progression builds on Turn Generative AI from an Existential Threat into a Competitive Advantage: off‑the‑shelf tools, proprietary data, and continuous feedback loops.
Initiate.
Doing existing work faster.
Most destination organizations today operate at the Initiate stage. Generative tools speed up existing tasks, and a person still reviews every output. An analyst uses AI to code open-ended survey responses. A marketing specialist lets AI suggest website optimizations or give feedback on content for generative engine optimization.
In this stage, data remains in silos and is governed informally. The primary change is the speed at which people complete tasks. Teams begin to understand where AI shines and where it struggles.
AI is useful and contained, but the underlying processes stay intact.
Research from GitHub found that among developers using GitHub Copilot, an AI coding assistant, completed tasks 55% faster than those who did not use it. The results were statistically significant. The developers were also less frustrated (59%) and focused on more satisfying work (74%).
The work itself is the same. The speed improved.
Internalize.
Shared knowledge surfaces.
At the Internalize stage, your own data is fed into the model through connected tools, knowledge bases, and curated sources. At times, the model is trained directly on your data, so there is shared organizational memory.
A knowledge assistant built from brand performance and partner records can answer specific questions about partner engagement tactics much more effectively than a generic model.
Teams start querying the same memory. As a result, silos soften.
This stage requires proprietary data, not just public tools or general models. Here, clean data becomes a necessity. It is the difference between an AI that learns from inconsistent, poor data and one that learns from high-quality data. As silos soften and data gets cleaner, your results improve.
You still steer, but AI carries more of the load as a partner.
Integrate.
AI handling end-to-end workflows.
In the Integrate stage, AI becomes the operating layer through which workflows take place. Data is clean and accessible. Through workflow setups, tasks like trend analysis, anomaly detection, demand forecasting, and partner-referral routing run continuously. It is also the most exciting phase, as organizations see how work gets done and really start to change with the workflow.
This stage reminds me of how old factories during the steam era adapted to the electric era. It is like the
shift from steam-powered, multi-story factories to modern, single-level production lines more than a century ago.
You redesign processes. Teams move into judgment-heavy roles, such as strategy, partner relationships, ethics, and governance. The AI runs, and your team supervises on an exception basis.
Here, governance shifts from a checkpoint to a continuous layer for an AI already in motion.
Interoperate.
AI becomes an ecosystem.
The horizon for organizations is Interoperate, where AI agents transact within and across organizations and sectors
as an ecosystem. Data is readable and flows across systems. Machine-readable standards allow agents representing travelers, hotels, restaurants, and attractions to negotiate and re-optimize in real time. Today, standards such as Google’s Agent-to-Agent protocol (A2A) and Anthropic’s Model Context Protocol (MCP) are beginning to define how agents discover one another, exchange capabilities, and coordinate tasks. Both are still in early stages.
Picture a family's travel-planning agent interacting with destination, brand, and peer-network agents to put together
a trip tapping into group chats, travel preferences, real-time pricing, and community characteristics. All negotiated across agents. The destination's brand, stewardship guidelines, and partner relationships are embedded in those protocols.
By this stage, clean data is table stakes, and governance is a strategic position in an ecosystem. Your role is to encode governance standards in shared protocols for stewardship, ethics, and brand integrity. And the exciting
part is continuously reimagining what becomes possible as AI capabilities improve.
Most destination organizations are at the Initiate and Internalize stages, with a handful planning for the Integrate stage.
Transformation efforts take years. The Destination Intelligence Arc attempts to describe that journey. Moving along the arc requires connecting AI initiatives to measurable progress, and laying the data foundation and mindsets at each stage. As generative AI makes everyone in your organization a data and AI user, someone needs to own the integrity of what the organization knows.
Here are some principles to consider on your own journey:
Summing it all up…
← scroll to see full table →
| Stage | Foundation | AI's role | People's role | What this looks like in an organization |
|---|---|---|---|---|
| Initiate | Off-the-shelf tools and chatbots like ChatGPT. Minimal customization. | Assists | Act | A research analyst uses AI to code open-ended survey responses. A marketing manager drafts campaign copy with an AI assistant. A sales coordinator summarizes a planner's email thread. |
| Internalize | Built on proprietary data. Specialized knowledge. | Partners | Steer | A destination knowledge assistant trained on the DMO's own visitor research and partner information. An AI co-analyst that drafts findings from the survey dashboard with local context. |
| Integrate | Always-on workflows. Data infrastructure as a system. | Runs | Orchestrate | Demand forecasting, anomaly detection on mobility signals, and partner-referral routing, all running continuously. A team supervising by exception. |
| Interoperate | An ecosystem of agents transacting across consumers and organizations. | Transacts | Govern | Agents transact with hotel, restaurant, transport, and experience provider agents to orchestrate and re-optimize visitor experiences in real time, against destination-level rules and standards set by people. |
“Data is not information, information is not knowledge, knowledge is not understanding, understanding is not wisdom.”
— Clifford Stoll
*********
