Surya Chikkala is Chief Digital & Technology Officer at Leadingresponse.
Most companies do not have an AI problem. They have a clarity problem. They know AI matters and are waiting for a cleaner moment. It never comes. What arrives instead is a competitor who stopped waiting.
I have spent enough years in technology to know the difference between adopting and absorbing AI. Adoption is what you can point to: a tool deployed, a workflow updated, a team trained. Absorption is when AI stops being a project and becomes how work happens.
To tell them apart, pressure-test the five layers: how people work, how you build, how you act on data, how clients experience what you build and how you scale.
The First Layer: How Your People Work
Before investing further in tools, ask where your people’s attention goes. Most organizations measure AI’s impact by time saved, but that is the wrong measure. The one that matters is decision confidence: Are teams reaching better conclusions, with less hedging, on cleaner data? If AI has made people faster without changing what they decide, you have accelerated old habits rather than changed how work happens.
I have seen this in work that once consumed a full day: pulling data, building the deck, drafting the narrative, all before anyone could begin the real conversation. When AI handles that preparation, the time collapses, but the more important shift happens in the room afterward. People arrive ready to pressure-test conclusions rather than having spent their energy building the container. Reaching that point takes more change management than leaders expect. A top-down rollout, where leaders model the behavior before their teams are asked to change, works better than a bottom-up one, though adoption will still stall in places.
The Second Layer: How You Build
The test for engineering is not whether AI accelerates the roadmap but whether it changes what you can build. If initiatives only compress timelines on plans you already had, you are optimizing rather than absorbing. The shift that matters is AI readiness becoming a design requirement instead of a feature added afterward.
I also want to be candid about what failed. One AI-assisted migration underperformed, not because the model was weak but because we had not defined its boundary conditions well. Give AI an imprecise objective, and it hands you confident output that falls apart under scrutiny; give it a precise problem, and it delivers.
That lesson governs how I scope every initiative. Where the approach did work, on a legacy system modernization, AI-assisted analysis surfaced complexity earlier than a traditional review would have, sparing remediation work that accumulates late in a migration.
The Third Layer: How You Act On Data
Having data was never the hard part for most organizations. Getting the right signal to the right person at the right moment is. Ask whether insights reach the decision-maker at the point of action or sit in a dashboard, waiting. Intelligence should be ambient, and the signal should find the decision.
I learned how much that principle matters when a signaling system flagged a campaign drifting toward a consent-language risk in a state with an especially demanding consent regime. The flag arrived before the campaign scaled, leaving time to correct the flow and relaunch in the same window. A weekly report would have meant cleaning up rather than preventing. In a regulated industry, that difference is reputational.
The Fourth Layer: How Clients Experience What You Build
Internal AI investment means little if it never reaches the people you serve. Ask whether clients can act on the intelligence you have built or whether it stays internal. For years, a gap has existed between what a lead generation platform knows and what the professionals it serves (financial advisors, attorneys and healthcare providers) can act on. Closing that gap becomes a priority because where retention follows demonstrable ROI, visibility is not a nice-to-have.
In practice, that means a live view of campaign delivery drawn from operational data: how a campaign is moving, where it is performing and where adjustment is warranted. Intelligence that once stayed internal becomes something the client acts on without hunting for it.
The Fifth Layer: How You Scale
Growth in a service business has always meant more people. That is a staffing plan rather than a business model, and eventually, the cost outruns the value. The question worth asking is whether AI agents are absorbing structured, repeatable work today or whether that remains an ambition.
When we mapped our account management workflow, most steps proved structured and rules-based, and judgment-intensive work was a smaller share than assumed. That imbalance is where AI earns its keep. We are building toward agents that handle high-volume orchestration while our people concentrate on work that needs them. Mapping comes first because an agent inherits whatever ambiguity the workflow leaves unresolved.
A Framework Any CDTO Can Use
Each layer carries a question worth asking honestly:
• On People: Has AI changed the quality of what teams decide or only their speed?
• On Building: Is AI changing what you ship or only how quickly?
• On Data: Do insights reach the decision-maker at the point of action or sit in a dashboard?
• On Clients: Can clients act on the intelligence you built, or does it remain internal?
• On Scale: Are agents absorbing structured work, or is that an ambition?
If your honest answer to most is the second option, you are adopting AI rather than absorbing it. That is not a criticism; it is where most organizations sit. But knowing which one you are is where every useful conversation starts.
What Absorption Requires
Getting from adoption to absorption takes three things most companies underinvest in: governance discipline, data infrastructure that is genuinely ready and leadership willing to commit before ROI is visible.
The companies that win will not be the ones that ran the most pilots but those that made AI structural, so intelligence and action live in the same place. Those five questions are ones I keep returning to, not as a checklist but as an honest measure of progress.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?