Why automation projects fail (and the simple framework to follow for success)

We had a call last week with a potential client who wanted to build an AI system that would automatically find prospects on LinkedIn, craft personalized messages, send connection requests, and follow up with a 7-touch sequence.

“I saw this guru do it and make $50K in a month,” he said.

When we asked what manual processes were eating up his time right now, he paused.

“Well… I spend about 6 hours a week copying data between systems. And probably another 4 hours sending the same onboarding emails to new clients. Oh, and I manually create invoices for every project.”

“So you want to automate something you've never done before,” I said, “but you’re not automating the 10 hours of manual work you’re already doing every week?”

Silence.

This is exactly why 90% of automation projects fail.

The 3 Types of Automation (And Why Order Matters)

After building 100+ automations, I’ve learned there are only 3 types of automation projects you can pursue.

Most people start with the wrong one.

Type 1: Automate What You’re Already Doing Manually

• Highest ROI
• Lowest risk
• Start here

Type 2: Automate Proven Strategies by Others

• Medium ROI
• Medium risk
• Do this second

Type 3: Experimental Automations

• ROI unknown
• High risk
• Only after you’ve mastered 1 and 2

Let me break down why this order matters…

Type 1: The No-Brainer Automations

These are the tasks you’re already doing manually that eat up your time every single week.

Examples:

• Moving data between systems
• Sending the same emails over and over
• Creating invoices and proposals
• Updating project statuses
• Following up with leads
• Onboarding new clients
• Content Creation

Why these win:

• You already know the process works (you’re doing it manually)
• You can measure exact time savings
• Zero risk of the automation “not working” - it’s just replacing manual steps
• Immediate ROI - every hour saved is money in your pocket

One client was manually creating project briefs in Google Docs, then copying information to their project management tool, then creating a Slack channel, then sending a welcome email.

45 minutes per new project.

We automated the entire flow. Now it takes 2 minutes.

At 8 projects per month, that’s 6 hours back in their schedule.
Every month.

Type 2: Copy What’s Working for Others

Once you’ve automated your manual processes, then you can look at proven strategies from other businesses.

Examples:

• Lead nurture sequences that competitors use
• Content distribution workflows from successful creators
• Client onboarding processes from similar companies
• Sales follow-up systems that have proven conversion rates

Why these work:

• Someone else has already validated the strategy
• You can see their results before you invest time
• Lower risk than experimental approaches
• You can often improve on their process

The key is finding strategies that are working in businesses similar to yours.

Don’t copy a B2C ecommerce funnel if you’re running a B2B consulting business.

Type 3: Experimental Automations

These are the “shiny object” automations that everyone wants to start with.

Examples:

• AI systems that prospect on LinkedIn automatically
• AI customer service bots
• AI avatar bots
• Complex multi-step sequences with unproven conversion rates

Why these usually fail:

• No validation that the underlying strategy works
• High complexity with unknown outcomes
• Easy to blame the automation when the strategy is flawed
• Lots of time invested with uncertain returns
• Development time is long and confusing, leading to higher costs

Here’s the thing - experimental automations can work. But only after you’ve built a foundation with Types 1 and 2.

Why Most People Get This Backwards

Type 3 automations are sexy.
They promise big results.
They’re what the gurus sell courses about.

Type 1 automations are boring. They’re just making your current life easier.

But here’s what happens when you start with Type 3:

The automation doesn’t generate the results you expected (because the strategy was unproven).

You blame the automation.
You think “automation doesn’t work for my business.”
You give up.

Meanwhile, you’re still spending 10+ hours per week on manual tasks that could be automated with 90% certainty of success.

The Smart Approach

1. Start with what you know works: your current manual processes.
2. Get 5-10 hours back in your schedule.
3. Use that extra time to research and implement proven strategies from others.
4. Then, and only then, experiment with unproven approaches.

By the time you get to experimental automations, you’ll have:

• More time to dedicate to them
• Experience building automations that work
• Confidence that automation can work for your business
• Systems in place to measure and optimize results

The Simple Test

Before building any automation, ask yourself: “Am I currently doing this manually, or am I hoping this will work?”

If it’s the latter, wait. Build the manual stuff first.

Your future self will thank you.

Talk soon,

Mitch

P.S. What manual processes are eating up your time right now? Hit reply and let me know. I might feature the most common ones in a future email (anonymously, of course).

Learn more about MakeBigMoves partner offers here: Partner Program | Whitelabel EOI.