Article Getting AI Unstuck: From Strategy to Scale
Why your AI initiative isn’t delivering and what to do about it
By Insight Editor / 3 Jun 2025 / Topics: Data and AI IT modernization IT optimization
You think you’re ready for AI. You’ve got a strategy. You’ve run the meetings. Maybe even launched a pilot. So why aren’t you seeing results?
The answer often lies in a misunderstanding of readiness. Many organizations leap into AI initiatives because they feel pressure to act. They want to stay competitive and not fall behind, so they move too fast.
They confuse having a strategy with being prepared. They jump into proofs of concept and pilot programs without fully understanding whether their data, systems, and teams are in sync. If they’re not in sync, things start to break down — projects stall, expectations go unmet, and AI gets blamed, even when the problem is upstream.
Readiness means alignment. It’s about knowing your data is in shape, your teams are on the same page, and your goals are tangible, measurable, and shared. When that groundwork is missing, even the best ideas will fall apart.
Where are organizations getting it wrong?
Most organizations work hastily. Leaders want to be on trend and show progress — especially eager board members. But without a data-first strategy and true cross-functional alignment, they’re just sprinting toward inefficiencies and poor outcomes.
AI initiatives often begin with good intentions but lack a strong foundation. The project might not align with what the business truly needs or is ready to support. Projects suddenly balloon in price because the organization didn’t pick the right model or set guardrails for the attempted adoption. And when it fails, the technology often takes the blame.
That lack of planning creates doubt; it makes the technology look expensive and hard to control. When that happens, decision-makers like CFOs get worried. They don’t see immediate value or ROI, so they start pulling back on scale and usage, which limits the tech’s long-term potential.
Too much investment in a poorly scoped or unsupported project can lead organizations to deprioritize AI altogether.
According to Businesswire, 68% of companies are experiencing internal friction due to AI adoption, and INC. reports that one in three executives says generative AI has been a massive disappointment.
That disconnect isn’t about the technology. It’s about the people, the process, and the preparation. AI adoption isn’t an IT or a business issue. It’s an organizational issue. Everyone — from security and legal to operations and executive leadership — needs a seat at the table. They need to understand their role, the outcomes they’re driving toward, and what alignment looks like.
AI adoption isn’t an IT or a business issue. It’s an organizational issue.
Insight for a course-correct
When teams jump into AI initiatives with a shaky foundation, even well-intentioned projects can drift off course. We’ve seen this firsthand.
One major healthcare provider had already begun developing predictive models to reduce patient length of stay — but without a clear business objective, the project lost momentum. That’s when Insight stepped in. We aligned the effort around a tangible outcome, brought the right stakeholders to the table, and implemented a monitoring system with clarified outcomes. The solution reduced ICU mortality by 19% and length of stay by 22% using a machine learning model that identified high-risk patients earlier.
In another case, a retail client had migrated data to the cloud and started testing generative AI. But they skipped the fine-tuning. Their models couldn’t interpret their internal product terminology. The result? Hallucinations and false outputs undermined confidence in the tools. We helped them retrain their models using the correct language and logic to produce reliable, usable results.
In both cases, Insight helped clients reconnect their ambitions to real-world outcomes. We don’t just plug in tools — we build the roadmap, align teams, and define what success looks like. That’s how we help AI deliver value, not just excitement.
What does AI readiness actually mean?
Bring struggling AI initiatives back on track with smart planning and a focus on measurable ROI.
What to consider when starting your journey
There’s no standard starting line for AI. Every organization enters the conversation from a different point, with its own priorities, capabilities, and constraints. That’s why one of the most valuable first steps is gathering a cross-functional group of IT, marketing, operations, HR, and other critical stakeholders to build alignment on what’s possible and meaningful.
From there, it becomes a matter of sorting signal from noise. Evaluate use cases based on both impact and feasibility. The sweet spot is something valuable enough to matter, but not so complex that it stalls out. That balance depends on your data readiness, tech stack, and organizational appetite for change.
Insight’s approach helps teams move from a wide field of ideas to a focused set of high priority use cases. Whether your data already lives in AWS or is spread across hybrid environments, we help you assess maturity, prioritize use cases, and start where you’re most likely to succeed.
And sometimes that starts with a technology bake-off.
Many teams have similar ideas, from AI solutions that will streamline internal workflows to ones that promise to enhance customer engagement. It pays to bring them all together to narrow the focus. Then test, compare, and validate those ideas so you can focus your efforts on where they’ll drive the most business value.
How to avoid false starts
Your AI initiative won’t take off if data lives in silos, stakeholders aren’t aligned, and governance and security protocols are an afterthought. Insight and AWS have the experience and expertise to set you up for success.
With AWS, organizations can quickly unify and centralize data using serverless tools like AWS Glue, making it easier to access and govern across systems. That includes structured and unstructured data stored on-premises, in SaaS tools, or fragmented cloud environments.
Once the data is in place, AWS offers tools like Bedrock for foundational models and SageMaker for custom development, allowing teams to fine-tune and build quickly without starting from scratch.
AWS also provides the architecture for responsible, scalable AI adoption. When you partner with Insight and AWS, you get more than infrastructure. You get guidance on governance, legal and security alignment, and stakeholder orchestration — often the difference between a pilot that delivers and one that gets shelved.
This collaborative model reduces risk and accelerates time to value. It ensures you’re laying the groundwork to make AI sustainable, secure, and impactful from day one.
Our proven approach to ensuring an AI pilot is successful
The move from planning to execution is delicate in any AI initiative. Treat the first week as a critical moment to translate vision into motion. That’s why we start with a structured kickoff, guided by a framework we call Radius. We’ve learned the best approach begins with bringing all the right stakeholders to the table. That means not just listing their names in a document but also confirming their roles, responsibilities, and time commitments.
Everyone in the room should understand two things: what’s expected of them, and how their work ladders up to broader business goals. Validate the roadmap. Confirm your communication plan. Align your technical and business teams. Because without that foundation, small missteps don’t stay small for long.
Next, we conduct a structured set of stakeholder interviews and assessments. These help everyone understand where the organization is in terms of data maturity, AWS cloud usage, automation practices, and governance. We closely examine existing data environments to identify silos, gaps, or legacy systems, and assess how to bring them together into a unified, accessible model that supports the use cases ahead.
Our AI approach is to create a repeatable framework that helps teams move with confidence. It includes maturity assessments, skills evaluations, and a clear view of the already established technical foundation. Week one is where we map that reality to your ambitions.
Equally important is to begin setting the rhythm of measurement and iteration. From the start, we decide how success will be tracked and how strategy might need to evolve. Your organization needs structure and flexibility to adapt as new insights emerge.
Strategy is the blueprint. Week one is when the building begins. And with the proper foundation, the rest of the project can move forward with clarity, accountability, and momentum.
The real cost of skipping ROI planning
When AI efforts fall short, it’s rarely because the technology failed. It’s because no one defined (or agreed) what success looked like from the start. This is why a third of senior executives say AI has underdelivered on expectations. That disappointment often stems from skipping the challenging but essential work of ROI planning.
Organizations invest significant time and resources into pilots and proof-of-concepts, technical teams spin up environments, business units dedicate time, and expectations are set. But without early value, those projects stall. One failed use case can cost hundreds of thousands of dollars. Multiply that across departments, and the risk adds up fast.
Accurate ROI planning doesn’t mean months of analysis. It means defining success at the outset and aligning teams on why this work matters: for the business, for the customer, and for the teams doing the work.
At Insight, we help close that gap with a focused engagement designed to assess where you are, define where you want to go, and identify the high-impact, realistic use cases that will get you there.
