To Use AI or Not? Let Me Tell You Three Stories

This past year, almost every week someone has asked me: "Should our company get on board with AI?"

I tell them: stop asking whether to get on board. First look at how others got on board — and how others didn't.

Today let me tell you three stories.

Two used AI, one didn't. And the results are all pretty interesting.

Story One: 100+ Posts a Day — Who Can Keep Up?

There's a TV network in South America. Big one. 114 TV channels, and their website pulls 11 million unique visitors a month.

Their pain point sounds almost mundane: they simply couldn't keep up with social media updates.

Think about it. On Facebook alone, they had to push out 100 to 150 pieces of content every single day. And it wasn't one person doing the posting — more than 20 people, scattered across departments, each running their own show, posting whatever they wanted.

Total chaos.

When to post? What to post? Who was making sure prime time wasn't drowning in low-quality content?

They did one thing: they let AI do the scheduling.

What did AI do?

• Analyzed historical data and predicted the best time to publish each post for maximum click-through rate.
• Automatically queued posts, delayed them, or bumped them up.
• Adjusted in real time as Facebook's algorithm shifted.

Four months later, the numbers came in.

• Daily clicks up 25%.
• Daily impressions up 61%.
• Facebook impressions up 52%.
• Twitter page views up 63%.

And — 14 hours of human labor saved every single day.

Here's the thing. What AI did wasn't write articles. It was "putting the right content in the right place at the right time."

That's something humans can't do — not a capability problem, a stamina problem. Two billion data points. You want a person to crunch those? They'd be at it till dawn and still not finish.

Story Two: Cookies Are Gone — How Do You Target Ads?

The second story stars a global software giant.

Third-party cookies are dying. Every marketer knows this.

Why does it matter? Because the old logic of ad targeting was "track user behavior" — this person browsed a car site yesterday, so I'll serve them auto insurance today.

Once cookies are gone, that playbook is dead.

So what do you do? This company picked a path: contextual advertising.

What's contextual advertising? It stops looking at "who you are" and only looks at "what you're reading right now."

You're reading an article about cloud computing — I'll serve you an ad for cloud services. Brutally simple.

But pulling this off takes AI. Why?

Because no human can scan hundreds of millions of pages across the web in real time and judge whether each article fits your ad. AI can.

It scans the text, images, video, and audio on a page, then makes a call on whether the page's "context" matches your ad.

The result? 61% of people who saw the ad took action, or got ready to take action. Ad viewability hit 93.2%. Overall brand awareness rose 4%.

Here's the key: zero cookies, zero personal data used.

The privacy era isn't a dead end. It's just a different track. And that track is called "contextual matching."

Story Three: We Just Don't Need AI

The third story is the most interesting. It's a counterexample.

There's an outdoor adventure travel platform — founded in 2019, already over 50,000 registered users.

What they do is deeply traditional: they have real humans write guides.

Not AI writing — real people, one article at a time. Which hiking trail is best, which mountain to visit in which season, which gear is actually worth it.

The founder said something I'll pass on to you directly:

"What users want is quality content. The internet is drowning in error-filled fluff pieces cranked out by content farms. AI looks cheap, but the quality just isn't there."

They just doubled down on SEO fundamentals — metadata, URL structure, alt tags, internal links, backlinks.

And then?

Their site speed couldn't keep up — and Google's algorithm punished them for it.

So they rebuilt the whole site. Ripped out the dynamic architecture, went fully static, stripped out the database calls, and added global edge nodes. Within a month, organic traffic was up 67% year over year.

Over 18 months, sales from organic traffic hit nearly $1 million. Five times growth year over year.

Paid ads? They tried those too. Turned out the conversion rate on paid channels was only a tenth of what organic delivered. Customer acquisition cost had climbed from $300 in 2019 to $1,000 now.

So they gutted most of their paid spend and bet everything on "great content written by humans + organic traffic."

And the bet paid off.

So What Do These Three Stories Tell You?

Let me line all three up side by side:

Case	Used AI?	Result
South American TV network	Yes — content scheduling	Impressions +61% in 4 months, 14 hours saved daily
Global software company	Yes — contextual advertising	61% of users took action, brand awareness +4%
Outdoor travel platform	No — doubled down on human-written content	~$1M in sales over 18 months, 5x growth

The ones who used AI won. The one who didn't use AI also won.

So is AI actually a must-have?

My judgment:

The key isn't "whether to use AI." It's "where you put your energy."

The TV network's pain point was scheduling. The software company's pain point was precise targeting after cookies disappeared. None of these were problems that "content quality" could solve — so AI was a huge help.

The travel platform's pain point was "users need to find a reliable hiking trail." The heart of that is content quality — so real humans writing won out.

AI solves problems of scale and matching. It cannot solve problems of insight and truth.

Which one is the core tension in your business?

That's a question only you can answer.

Once you've thought it through, you'll know whether you should get on board with AI.