HTEC’s Post

Asitha Goonewardena Exactly. The blame-the-technology pattern is frustrating because it masks the real issue - insufficient discovery work upfront. What I've found helps: treating data readiness as a gate, not a checkbox. If the assessment reveals data gaps, that becomes the priority. AI delivers value once the foundation is solid. Saves a lot of wasted effort and disappointment down the line.

Data quality is foundational to AI value. Most organizations pick AI tools before understanding if their data can support the use case. If your data governance needs work or your architecture wasn't built for the questions AI needs to answer, results will fall short. The organizations getting ROI treated data strategy as a prerequisite, not parallel work. They mapped where data lives, identified quality issues, and closed gaps first. Strong data foundations enable AI value. Without them, even good AI struggles to deliver.

We're dealing with a fact that, due to the human like capabilities of AI, all of us are attaching human traits to this new technology. So trust has become a metric. It has a lot to do with probabilistic output of AI, which we are not used to, as we've been using deterministic systems so far. In order to build trust, we need to start with clarity.

You can start small, to prove the value. Keep people involved. By these step simple things may emerge which could grow into actual ROIs, and celebrate the small steps.

Some thoughts on ROI thinking. Financial diemnsion is the most obvious one (cost optimization, new revenue, etc.) but operational and startegic are equally important. I often hear the need to optimize time-to-decision and treating it as driving metric behind investment decisions. Competitive positioning is another common topic that surfaces in different conversations in this context.

Boris Paunović I'd even extend the data readiness beyond data only to the entire organzational context. You may have data in a solid state but there may be other obstacles that will still prevent successful AI implementation. Think operations, skillset, infrastructure, leadership mindset, alignemnt with overall goals, etc.

Asitha Goonewardena there is one interesting irony here, as AI has also made it much easier for organization to deal with the issue. It both emphasizes the problem and a tool to solve it.

We are in Halloween season, best time for the scariest topic: Data Governance.

Bastian Rodrigo Keep asking the lighthouse questions at every phase. It's ok to fail, but fail fast. Where fatigue and waste happens is trying to keep the experiments alive. ROI gives you a good focal point. If your initial estimate is not working out, being realistic is the key. Great way to build organizational muscle is to start on smaller internal projects, though lower in ROI, they also have a lower risk, and faster turnaround.

Boris Paunović A Common issue I have come across in many AI project. Without the proper Data availability and plan, organizations jump for AI solutions and eventually loosing the ROI and blame technology.