The danger of automating everything and the "90% Rule"«

199

    We live in the golden age of "do it once, run it a thousand times." Tools like Make, Zapier, and the ubiquitous AI sell us a seductive dream: Set up a system and disappear while the money (or traffic) comes in on its own.

    But I have bad news: Automation at 100% is a trap.

    If you delegate judgment to a machine, you're not scaling your business; you're scaling your mistakes. And worse, you're eliminating the only thing that sets you apart from a Python script: your humanity and your judgment.

    Today we're going to talk about why obsessing over "autopilot" is dangerous and why you should adopt the philosophy of the automatic 90%.

    The Set and Forget fallacy«

    The biggest productivity mistake I see today is treating automation as an "off" switch for the brain.

    When you automate a process completely without supervision (especially if it involves customer communication or content creation), you expose yourself to three fatal risks:

    1. The Uncanny Valley: Your emails or posts sound almost Humans, but something's off. A phrase out of place, a lack of context. The reader notices, feels repulsed, and your personal brand plummets.
    2. Multiplication of errors: If you make a mistake in a manual process, you affect one customer. If you make a mistake in an automation loop, you can send the wrong email to 5,000 people in 30 seconds.
    3. Loss of Feedback: By not engaging with the process, you lose sight of it. You lose touch with what your customer truly needs today because you're basing your business on rules you set up six months ago.

    The 90% Automatic Rule (Human-in-the-loop)

    The solution is not to return to manual labor and breaking rocks. The solution is the 90/10 methodology.

    The premise is simple: Let the robots do the heavy lifting (collecting data, formatting, classifying, drafting), but never Don't let them post or submit without a human adding the final 10%.

    That 10% is where the value lies. It's quality control. It's the "soul".

    What does this look like in practice?

    Let's bring it down to earth with real-world examples of how to apply this automatic 90% vs. the disaster of the 100%.

    1. Creation of Content and SEO

    • Approach 100% (Dangerous): You connect RSS to ChatGPT and ChatGPT to WordPress. Five articles are published automatically each day.
      • Result: Google penalizes you for junk content, you blow up data, and your audience realizes there's nobody at the wheel.
    • 90% Approach (Winner): The AI looks for trends, generates the structure and writes a first draft (90%). You come in, correct the tone, add a real personal anecdote and verify the data (10%).
      • Result: You publish three times as fast, but with author quality.

    2. Prospecting on LinkedIn or Email (Outreach)

    • Focus 100% (Spam): A bot sends generic messages to anyone who meets a filter. "Hi {Name}, I love your profile…".
      • Result: They block you, you burn your domain, and you look desperate.
    • 90% Approach (Effective): The system identifies the lead, extracts their latest posts, and drafts a message based on their recent activity. You read the draft, adjust a phrase to make it sound genuine, and click "Send.".

    The final 10% is your competitive advantage

    In a world where everyone has access to the same tools (GPT-4, Claude, Make), the automation It's a commodity. It is no longer a competitive advantage in itself.

    What cannot be automated is:

    • Real empathy.
    • The cultural context of the time.
    • Subtle humor.
    • The long-term strategy.

    If you automate everything, you become a commodity. If you use automation to remove the "fat" from your work and dedicate your energy to that final 10%, you become a sniper.

    Conclusion: Be the pilot, not the passenger

    Technology should be your exoskeleton. It should allow you to run faster and lift heavier, but You must decide where to step.

    Review your workflows today. If you have processes running without you (or anyone on your team) ever looking at them, be careful. They're probably slowly damaging your brand.

    Find that sweet spot: 90% machine for efficiency, 10% human for excellence.

    ✉ SHALL WE TALK?

    Let's talk

    If we share interests in artificial intelligence, data, or software development, I'd be happy to exchange ideas. You can also contact me for professional matters.

    Read the blog