The Monday Morning AI Test

Tomorrow morning, ask 5 employees what wastes their time. That's where your AI should start. Not another pilot. A real solution.

Hi there,

Your AI pilot worked great. Three months later, nobody’s using it.

Research shows 74% of companies can’t scale AI beyond pilots. Only 4% get real value. Meanwhile, that 4% are doing the same work you do, just better, faster, and at half the cost.

The problem isn’t the technology. It’s that pilots get built for boardrooms, not for people doing actual work.

Start With What Actually Wastes Time

Tomorrow morning, find five people who do frontline work. Ask them: “What ate most of your time last Monday?”

Write down their answers. Circle whatever shows up most. Watch someone do that task the normal way. Time them. See them jump between systems, copy data, struggle through it.

Then have them try the same task using ChatGPT or Claude. Time that too.

If AI doesn’t save at least half the time, pick a different task. If it does, show both ways to the team. When three or more people say “I want the AI way,” you’ve found your starting point.

That’s it. One day. No consultants. No roadmap. And if it works, you’ve just found a way to do that task twice as fast at zero additional cost.

Why This Simple Test Beats Six-Month Pilots

Only 31% of businesses have scaled AI to production according to KPMG. McKinsey reports over 80% see zero earnings from their AI efforts.

But look at the ones who succeed. JPMorgan got 10-20% efficiency gains on software development. Walmart cut planning time by 67% and now handles 3 million daily queries. UnitedHealth runs approximately 1,000 AI applications in production. Better quality, faster delivery, lower cost.

They succeed because they spend money differently. BCG’s research found winners invest 70% in people and process, 30% in tech and data. Most companies do the opposite.

Your team is more ready than you think. McKinsey found leaders believe only 4% of employees use AI for real work, but employees report 12%. That’s three times higher. They’re already using ChatGPT at home. They need permission to use it at work.

Test Your Foundation First

Pilots work because everything’s perfect. Clean data, volunteers, attention. Real life isn’t perfect.

Before starting, answer these questions:

• What data actually exists?

• Where does it live?

• How often does it update?

• What percentage works?

• Who owns it?

Can’t answer? Stop. Fix that first. You’re building on sand.

Build Your Team of Five

You need five specific people:

1. Translator - Someone who explains tech to non-tech people (usually in sales or customer service)

2. Useful Skeptic - Your doubter who’ll actually try it fairly

3. Data Person - Someone who knows where data really lives (usually in operations)

4. Process Owner - Someone who owns a broken process and can fix it

5. Shield - A senior person who blocks the “we’ve always done it” crowd

Exclude anyone who wants committees or six-month plans. They kill progress.

Start Small and Copy What Works

Walmart started with one planning task. Just one. They cut the time by 67%. Then they expanded. Now they handle 3 million queries daily.

JPMorgan focused on developer pain points. They got 10-20% efficiency gains. Then they expanded.

UnitedHealth kept proving “impossible” tasks were possible. Now they run about 1,000 AI applications.

The pattern is clear: start with one real problem, prove it works, then replicate. No new tools needed. Use what you have.

Address Real Concerns Directly

IBM research shows 35% of workers will need reskilling in the next three years. People are right to be concerned. Address it directly:

When they say “AI will take my job,” show them what new work gets created. Document the training plan.

When they say “I’m not technical,” show them that knowing the business beats knowing code. They need simple instructions, not manuals.

When they say “This reduces my value,” make them the quality controllers. They make decisions, AI does repetitive work.

Make Your Choice Tomorrow

The Monday Morning Test takes one day. Most pilots take six months. Both reveal the same thing: people waste time on tasks AI could handle better.

While you read this, competitors are delivering more with the same headcount. While you plan, they’re shipping. While you debate, they’re improving.

The 4% of companies succeeding with AI aren’t special. They just started fixing real problems instead of running pilots. Every task they improve is done better, faster, cheaper.

Tomorrow you can either run this test or schedule another meeting about AI strategy.

Need help? The test is simple, but every organisation has unique barriers. I can help you identify yours and build a practical AI change plan. Let’s turn your test into your first win.

Regards,

Brennan