Gartner found that around 85% of AI projects don’t live up to their early promises, and there’s often a gap between what a business needs and what employees want.  A 2025 Prosci survey said 63% of companies blame people issues for poor AI adoption—such as concerns about job loss or lack of understanding of the technology.  Another poll of U.S. workers showed 71% are uneasy about AI, and three-quarters fear it could make jobs vanish.  Many employees also worry about pay cuts or falling behind if they don’t learn new AI skills.

These feelings get worse when AI tools are dropped on teams without including them in the process. 

Seems strange you're reading this on an AI agency's website, right? Haha, stick with me.

There's a missing factor here, though. Having led change for many years, there's a simple framework I follow. In short: It's not about features, it's about people.

The People‑First Approach

At Acadia AI, I always remind myself that AI should help people, not replace them.  Studies on human‑centered AI show that trust grows when people understand what a system does and feel it works for them.  We invite users into the design process, hold workshops that mix engineers with the everyday people who will use the tools, and share how the system enables the teams to still make decisions, create, and build.  These sessions make sure the tools solve real problems and get everyone excited to use them.

We also know that emotions matter.  EY’s research found that 80% of workers would feel better about AI if they received training and upskilling.  So we offer hands‑on practice and explain the rules that guide the technology. This people‑first approach turns fear into confidence.

The 3S Model: Simplify, Support, Scale

To make our philosophy easy to follow, we created a simple framework called the 3S Model.  It has three steps: Simplify, Support, and Scale.

Simplify

Don’t try to change everything all at once. One of the biggest mistakes in AI adoption is taking on too much at once. The best teams focus on one repetitive, time-consuming task that AI can handle well — for example, drafting reports, summarizing interviews, or tagging data. Before AI, summarizing 20 interviews might take 10 hours; with AI, you can get a draft in 20 minutes and refine it in a couple of hours. Those quick wins demonstrate to the team that AI is useful.

Simplifying also means setting clear boundaries. Decide what the AI handles (like summarizing or identifying patterns) and what people do (like reviewing work and adding context). This maintains AI as a co-worker, not a boss.

Support

Once you achieve a small success, support your team so they can grow with the technology. Prosci’s research shows that lack of training causes 38% of AI adoption issues, and people seek clear guidance and transparency. Our support plan includes:

• Training and upskilling: Workshops, guides, and one-on-one coaching to ensure everyone feels capable.

• Transparent communication: We explain how the AI functions, what data it uses, and what limits we impose. When users understand the “why,” they trust the system more.

• Feedback loops: AI needs to learn from us. We encourage feedback, corrections, and new ideas, helping the system improve over time.

• Leadership alignment: Nearly 43% of AI failures occur when leaders fail to provide enough support. We ensure leaders link AI projects to business goals and clearly communicate the vision to everyone. Communication and clarity is kindness.

Scale

After you’ve simplified and supported your first project, it’s time to scale wisely.  Scaling doesn’t mean using AI everywhere; it means using it where it makes sense and complements existing work.  Real‑world success stories share common themes: they solve clear problems, use good data, keep people involved, and track results.  Our scaling rules are:

1. Solve real problems: Only expand to new uses when there’s a clear, measurable need.

2. Prepare your data: Great data is the fuel for smart AI. We set up data standards and ethical practices to prevent bias.

3. Augment, don’t replace: AI handles routine tasks, while people retain judgment calls.

4. Measure and adjust: We track both short-term wins like time saved and long-term impacts such as quality and cost savings, following guidance that states successful AI programs align strategy and user experience and measure dual outcomes.

How ADKAR Fits In: Guiding Individual Change

Even with a simple 3S framework, we still need to help each person navigate through change. That’s where Prosci’s ADKAR Model comes in. I've been using this framework for years, teaching leaders I manage about these steps, and I’ve always seen amazing results. ADKAR stands for Awareness, Desire, Knowledge, Ability, and Reinforcement, and focuses on individual change. Prosci notes that organizational change succeeds only when you prepare, equip, and support people throughout the process. Here’s how ADKAR complements the 3S Model:

• Awareness: Before using AI, people need to know why the change is happening.  Leaders should share a clear reason and show how AI fits the organization’s strategy.  Without awareness, employees may see AI as a threat.

• Desire: People must want to participate in the change.  Show them how AI will help them succeed, involve them in planning, and provide chances for hands-on learning.

• Knowledge: Training provides people with the information and skills needed to use AI effectively.  Structured learning paths and mentorship build confidence.

• Ability: Knowledge isn’t enough; people need practice and support to apply what they’ve learned. Coaching, peer collaboration, and real-world projects build ability.

• Reinforcement: Finally, organizations must recognize and reward new behaviors to make the change stick. Encourage experimentation and provide safe spaces to try AI.

When combined with Simplify, Support, and Scale, ADKAR ensures that no one gets left behind.  It turns our people‑first philosophy into a step‑by‑step journey for every team member.

A Real‑World Story: CarMax Shows It’s Possible

Let’s look at CarMax, the used‑car retailer, to see how this works.  CarMax teamed up with OpenAI and Microsoft to build a system that summarized over 100,000 customer reviews into about 5,000 easy‑to‑read highlights.  If people tried to do this by hand, it would take 11 years, but the AI finished in a few months.  This is a perfect “Simplify” example: they chose one big, repetitive task and let AI handle the first draft so employees could focus on refining the highlights.

The project also shows the importance of Support.  Because the AI did the heavy lifting, CarMax’s team could focus on creative work like deeper content and better customer communication.  Leaders made sure everyone knew how the tool worked and how to check its results.

CarMax’s success also shows how to Scale.  By tracking results (faster content creation, better website search, and happier customers), the company gained confidence to use AI in other parts of the business. They didn’t cut jobs; they redeployed people to higher‑value tasks. Similar stories come from different companies: Walmart saved $75 million and cut 72 million pounds of CO₂ by using AI to optimize truck routes, and JPMorgan’s COIN system freed up 360,000 staff hours by automating document review. These cases all prove that AI works best when it solves a specific problem, has good data, keeps humans involved, and measures success.

A Hopeful Finish: AI Can Make Us Better

I believe AI should enhance human creativity, not diminish it. When AI takes care of repetitive tasks, we gain more time to ask better questions, explore deeper insights, and make smarter decisions. People‑centered AI also keeps us safe: clear rules and ethical guidelines build trust, and ongoing feedback ensures the system remains helpful.

By Simplifying our first steps, Supporting our people and Scaling thoughtfully, we can bring AI into our work without feeling overwhelmed.  The goal isn’t to cut jobs—it’s to help people do their best work.  When AI serves people, everyone wins.

Sources // resources from my AI agents

• AI Adoption: Driving Change With a People‑First Approach – Prosci https://www.prosci.com/blog/ai-adoption

• Human Centered AI: Principles, Benefits, Challenges, and Industry Examples – Tredence https://www.tredence.com/blog/human-centered-ai

• AI Adoption That Works: 5 Enterprise Case Studies – NineTwoThree https://www.ninetwothree.co/blog/ai-adoption-case-studies

• The Prosci ADKAR Model – Prosci https://www.prosci.com/methodology/adkar