[PRIMO Tech-a-Break] From the Creator of Nong Baitoey: Before You Believe AI Is the Answer

[PRIMO Tech-a-Break] From the Creator of Nong Baitoey: Before You Believe AI Is the Answer

These days, everyone is talking about AI in Martech.
But as someone who helped create it, I want to say this:
Don’t jump on the bandwagon just yet.
If you believe that AI CRM is the answer to everything — I can tell you now, it’s not.
AI doesn’t work just because it’s AI — it works because the data structure and systems behind it are truly ready.

Many of you may have seen that PRIMO recently launched “Nong Baitoey,” our AI built into the PRIMO Loyalty CRM.
Several of our clients are already using it.
Let me take a moment to say — as someone who helped build Nong Baitoey,
I’m proud that we’ve made it truly usable in Thailand’s CRM market — and we’re proud to be the first to launch at this speed.
Let me share the real reason we were able to launch before anyone else.

It’s not because we had a more advanced model —
it’s because we laid the groundwork from the very beginning.
And that’s why I said earlier: don’t be too quick to believe that AI will magically solve everything.

There are many other factors that need to be in place.
You need to be truly ready before you can unlock its full potential.

1. Garbage In = Garbage Out

  • AI doesn’t fix bad data. If you feed it garbage, it gives you garbage (Garbage in, garbage out).

  • Some CRM systems collect data with no clear formatting rules — duplicate names, inconsistent phone numbers, broken field mappings.

  • The result? AI runs calculations on messy data, and the insights it produces simply can’t be trusted.

2. AI Makes Problems Appear Faster

  • Many CRM systems using AI for segmentation discover that the “new customers” group includes people who actually bought from them 3 years ago.

  • Hidden issues don’t get solved — they get magnified.

  • Sometimes, AI becomes more like a spotlight, exposing the chaos in your data.

3. AI Doesn’t Know Your Business Logic — Until You Teach It

  • Cross-brand rewards, mission-based campaigns, tier-based structures — every brand has its own model.
    If you don’t clearly define and input your business logic, AI will give you insights that don’t work.

  • Many systems rely on generic, default models that can’t understand your specific context, making the insights unusable.

  • There’s a big difference between “fine-tuning” and messy data with no pattern.
    If your data is structured and consistent, it doesn’t take long to train AI to understand your business — and the insights you get will be far more powerful.

4. AI Is a Tool — Not a Human Replacement

  • AI is not a silver bullet for bad data. Again — garbage in, garbage out.

  • Without consistent formatting rules, CRM systems collect unclean data — duplicated names, misspelled phone numbers, broken field mappings.

  • The outcome? AI ends up working on unreliable data, producing insights you simply shouldn’t act on.

Straight to the point:

Before using AI in Martech, make sure:

  1. Your data is truly ready

  2. Your business logic is embedded

  3. Your team knows how to use AI

AI is not a magic button.
It’s a powerful engine — but if the road is broken and the driver isn’t ready, you won’t get anywhere.

 

And the reason PRIMO was able to build and launch AI faster — and more effectively — than many others is simple:

Our data structure was designed from day one with marketing in mind.
It’s complete, well-structured, and built with a strong foundation. So when AI was introduced, it integrated seamlessly — no need to tear anything down and rebuild.

Our data architecture was built in a way that AI can understand naturally, making it ready for real-world use from the start.