Each major AI has its own strength.
Grok is plugged into X, so it’s quick to pick up what people are talking about right now. Gemini sits inside Google’s world of search, Maps and Workspace. Claude is known for working through long documents and careful reasoning. ChatGPT is the all-rounder most people try first.
Pick any of them and you’ll get smarter marketing ideas than you had yesterday.
But there is one place none of them look unless you take them there: your profit.
The question AI rarely asks
Ask an AI to improve your marketing and it will talk about reach, engagement, content ideas and ad copy. Useful things.
It rarely starts with the question that decides whether your marketing is working: “How much does it cost you to get one customer, and how much is that customer worth?”
That question lives in your sales records, your accounting software and your ad accounts, scattered across three or four places that no AI can see until someone connects them.
Three numbers that matter more than likes
1. Customer acquisition cost (CAC). Total marketing and sales spend divided by the number of new customers. If you spent 50 million VND last month and won 25 customers, each one cost you 2 million.
2. Customer lifetime value (LTV). The profit a customer brings over the whole relationship, not just the first purchase. A customer who spends 1 million VND per visit at a 40% margin and returns 10 times is worth 4 million VND in profit.
3. Break-even ROAS. How much revenue your ads must return for every dong spent before you stop losing money. With a 40% margin, you need at least 2.5 VND of revenue for every 1 VND of ad spend just to break even.
In the example above, a 2 million VND customer worth 4 million in profit is a healthy business. But if that customer only buys once, the same campaign loses money on every sale. And the dashboard can look great while the bank account shrinks.
(These figures are an illustration, not a real client.)
Why AI can’t do this alone
It’s not that AI can’t calculate. It’s very good at calculating. The problem is the inputs:
- Your margin per product is in your accounting system.
- Your repeat-purchase rate is in your sales records, if anyone is tracking it.
- Your ad spend is split across Facebook, Google and TikTok, each counting conversions its own way.
- Your offline sales from phone calls and Zalo are often not connected to any of it.
Until someone gathers these numbers, cleans them and agrees on definitions, any AI is optimising a business it cannot see.
Likes don’t pay salaries
The most dangerous marketing report is one full of growing numbers that aren’t money: impressions, followers, clicks, views. They feel like progress. They can all go up while profit goes down.
Marketing that cannot be connected to profit isn’t an investment. It’s an expense with a nice dashboard.
Where DMG fits
At DMG, we measure money, not vanity metrics. We connect ad spend, sales and margin so every campaign answers one question: did this make the business more profitable? It’s step four of our DMG Academy framework, net revenue optimisation, tracking CAC against LTV and break-even ROAS. AI helps us analyse faster. The numbers keep everyone honest.
Not sure what one customer really costs you? Let’s work it out together. Talk to DMG.
