68% Are Already On Board. What Are You Waiting For?

A while back, I came across a survey of marketers in Bangladesh, and it gave me a jolt.

120 marketing practitioners, 12 senior marketing executives. Questionnaires plus interviews. One single question: do you use generative AI in your marketing? And how's it going?

Turns out, 68% of companies are already using it.

68%. In Bangladesh.

You might find that odd: isn't this a rich-country game? What does it have to do with a small business owner in Dhaka?

That's what I thought at first. But think about it: Amazon's recommendations, Netflix's watchlists — the things AI does most brilliantly — all come down to the same thing: using data to guess what you like, then putting the right content in front of you.

Guessing what you like — is there any difference between a consumer in Dhaka and a consumer in New York?

No.

The demand is identical. What differs is only how soon the tools reach your hands.

What Are They Using AI For?

The survey listed a bunch of use cases. Let me walk you through them.

45% of companies use it to generate content automatically. Copy, images, posts — the machine produces the first draft, humans edit.

35% use it for personalized email. It used to be one mass blast; now it's a personal message for each customer.

30% have deployed AI chatbots. A customer asks at three in the morning, the bot replies in a second.

And another 25% use AI to predict what users will do next.

Have you noticed? These are all the most labor-intensive tasks in marketing.

A ten-person marketing team can be drained dry just writing copy, answering messages, and sending emails every day. Now that AI has taken these over, what do the humans do? They make judgment calls.

And the results? Companies using AI saw customer engagement rise by an average of 23%. Marketing ROI is trending up as well.

23%. Doesn't sound like much?

Think about it: a small company with a limited budget and limited headcount, no overtime, no new hires, and engagement up by twenty percent. Run the numbers any way you like — it pays off.

But There's Always a Flip Side.

The survey also asked another question: for those not adopting AI, why not?

The number one answer, chosen by 45% of companies, was the same: too expensive.

The two runners-up: can't hire anyone who knows how to use it, and data privacy is murky.

These three walls, honestly, each one is harder to tear down than the last.

Cost is a money problem. The good tools charge monthly, priced in dollars. A Bangladeshi SME's annual marketing budget might be less than a single quarter of their subscription fee.

Talent is a people problem. You buy the AI tools, but nobody can do prompting, nobody understands data — the purchase just sits there collecting dust.

Privacy is a rules problem. Can you collect the data? Store it? Feed it to a model? When the rules are vague, companies don't dare move. Nobody wants to wake up one day and find they've crossed a red line.

So where emerging markets get stuck is never the technology — it's the money, the people, and the rules around the technology.

So What Now?

The researchers offered a few recommendations, and I think they hold up anywhere.

For companies: don't start by buying the most expensive full suite. Start with the single most labor-intensive task — say, wiring copy generation into your workflow first — and once you see results, go further.

For governments: make the data rules clear. Vagueness is scarier than strictness — with clear rules, companies know which way to go.

For the industry: train people. Tools will keep getting cheaper; people who know how to use them won't multiply on their own.

To close, I want to come back to that 68%.

Here's the interesting part: the fewer resources you have, the bigger AI's leverage. Big companies have plenty of staff — for them, AI is icing on the cake. Small companies have none — for them, AI is a lifeline.

Whoever hands off their repetitive work first gets to spend the saved time on the things machines can't do.

What can't machines do?

Understand your customer, as a person.

Run the numbers yourself.