Using AI for Digital Marketing: What It’s Actually Good For, from Real Hands-On Experience

This is AI for digital marketing from someone who actually uses it. AI now plays a bigger role at work than ever, especially for people in marketing. At Digital Break Time, I have brought AI into my own work too — and anyone working at a digital agency, as a freelancer or as a solopreneur (a one-person business) can apply the same ideas.

AI for Digital Marketers Is an Assistant, Not Someone Who Runs Your Ads for You

With a heading like this, many people might assume I let AI create campaigns or adjust ads all by itself. Let me be clear upfront: I don’t go that far — not 100%.

From the point of view of someone who does digital marketing and runs ads personally, AI is there to support me and assist with the tasks that eat up a lot of time, so I can get the real digital marketing work done faster. Examples include making reports that are easy to read, setting up small automations, and creative work such as images, videos and ad copy. You don’t need to be a “tech AI bro” to do any of this.

1. Use AI to Build Dashboard Reports That Are Easy to Read — for You and Your Clients

Traditional reports — pasting in lots of numbers, or taking screenshots and explaining them — have a big problem: the text is tiny, there is a lot of data, and sometimes clients simply don’t understand it. I used to make reports exactly like this, and it made communicating with clients quite difficult.

Some people do have the skills to make better-looking reports with tools like Power BI or Looker Studio. But these tools take fairly advanced skills, and you have to filter the data first to decide what to use and what to leave out. In the worst case, you might even need to hire a data visualization specialist just to get a report that is easy to read, looks good and can be reused.

AI let me skip almost all of those steps. I just export the report I need and tell the AI which metrics to focus on (in the past, I had to clean the data myself). It can also combine data from other platforms. Then I tell it what kind of data visualization I want — bar charts, line charts or something else — and that’s it: I get a good-looking report. If I want to keep using it in the following months, I also ask it to add a spot for uploading new files.

Reporting used to take a long time for each client. With AI doing the heavy lifting, it’s now easy, convenient and fast — just a little over ten minutes.

Of course, I still have to check that the numbers match the real report. That part is the human’s job: checking accuracy and polish so the report stays trustworthy.

2. Let AI Create Daily Reports Automatically Instead of Downloading Data Every Day

Before I knew about AI, downloading data and making daily reports was very boring, because it’s the same repetitive task every single day. If you only have a few clients, you can download the data and do it yourself. Back when I worked at an agency, we subscribed to Supermetrics or Zapier to pull data automatically. That made daily reporting smooth, but it came at a significant cost.

Daily reports don’t have to be boring anymore once you use AI — especially an AI agent, which is AI that can make decisions and take actions on its own (personally, I currently use Claude Cowork). You can set a schedule for the days and times you want it to run. And with a browser extension like Claude in Chrome, the AI can access the web pages you have open, download simple report files and finish the daily report in very little time. It’s a one-time setup that isn’t hard and delivers real results: you save time, you don’t have to download anything yourself, the output is the same, and you save money too.

One caveat: if you’re pulling large amounts of data from many platforms, Supermetrics or Zapier are still the better value and the more sensible choice. They work through APIs (direct connections between systems), so they are more accurate — and they can also work together with AI agents like Claude.

3. Use AI to Analyze Ad Creatives: Which Ones Work, Which Don’t, and Which to Keep Running

Let’s be honest: AI is far better than us at handling large volumes of data. In an ad account with many campaigns and lots of creatives — images, videos and countless versions — it’s hard for a person to look at everything. AI can, because it covers every level: campaign, ad set, ad and even placement.

Here’s a real example. At first, I simply asked the AI which ads should keep running and which should stop, for a messaging campaign aimed at generating leads (potential customers who share their contact details). As expected, its analysis matched my own view: ads that brought in lots of messages but few leads, or ads that spent a lot without getting any messages or leads, should be turned off.

What was really new was this: the AI suggested I put the price shown in each creative into the ad’s name, so we could test how much showing a starting price for these products and services actually mattered. I did that the following month and then analyzed the results. Ads that showed a price did have a higher cost per message — but a higher lead rate, because people who clicked to ask were already fine with the price, so they turned into leads more easily. Ads without a price got more messages, but a much lower rate of turning into leads. That made me realize that in some areas I hadn’t thought of, AI can analyze things more thoroughly than I can.

4. AI Can Help Find the Cause When Your Numbers Look Abnormal

Have you ever seen ad results that just look wrong — some numbers unusually high, others below average — even though you didn’t change anything suddenly or do anything special? You then have to dig through all kinds of data to work out why a number rose or fell so sharply. Tracking down the cause can take a lot of time.

A real example: Google Ads conversions for a client’s website dropped abnormally. There were some conversions in the first half of the month, but almost none in the second half. In the past, I would have had to hunt down the cause myself. With AI, I got several hypotheses, which you can ask it to list point by point. One of them was that the Google Tag Manager tracking (the code that records conversions on the website) had dropped off, because the client had updated their website at some point and Google Tag Manager hadn’t been reinstalled. The fix was easy: add the Google Tag Manager base code back to the website. It took hardly any time.

5. AI Makes Writing and Video Work Much Easier

Most of the topics and ideas on the Digital Break Time website still come from me. What AI does is expand those ideas, making them broader and deeper. Take this article: I came up with the topic myself, with an introduction and about 3–4 subheadings. Then I asked AI to help expand the ideas, and it added more subheadings. In this article, the sections AI added are:

• AI can help find the cause when your numbers look abnormal
• What I don’t use AI for (the full idea is in the next section)

Everything else is my own. As you can see, AI fills in quite a lot, making the content deeper, broader and more complete. But all the details are still written by me, a human, because Google still wants E-E-A-T content (Experience, Expertise, Authoritativeness and Trustworthiness). This article focuses mainly on experience and expertise — exactly what Google Search and AI Overviews look for.

As for video, Digital Break Time doesn’t do it often (though we do take on video work, mainly small projects: sales-focused videos made specifically for running ads). When a job comes in, I use AI to help write the script and lay out the storyboard, which makes it much easier to picture the final video. I can now handle all of this on my own, when it used to take several people.

6. What I Don’t Use AI For

Even though AI does a lot for me, I currently have a few rules about things I never let AI do, such as:

• Letting AI create campaigns, adjust budgets or turn ads on and off entirely by itself, without me deciding
  Mistakes here cost real money that you can’t get back. So I let AI review campaigns, but a human makes the decision. Humans also understand the context of the business, and there may be factors outside the ad account that AI doesn’t know about. That’s why it’s still better for a human to handle these steps.
• Sending reports to clients without checking them first
  AI is accurate and fast, but every AI answer usually comes with a small note saying that AI can make mistakes — so you always need to check. AI may produce a beautiful report, but you must always verify the data and numbers first. Doing this, I have found errors many times: wrong numbers, incorrectly combined results, or metrics pulled in that I didn’t ask for. Reviewing and correcting is simply part of using AI.

7. Summary: Where Should You Start Using AI for Digital Marketing?

If you want to start using AI for digital marketing, begin with the tasks that are the most repetitive and take the most time — you can start right away. For me, that was reporting, because it’s a regular task that takes a lot of time. Then gradually hand over more of the tasks you think suit AI. As this article shows, you can also use it for other things, such as small automations, expanding ideas for writing, writing scripts and getting more well-rounded analysis.

Many people wonder whether AI will replace marketers. My personal view: repetitive tasks will be replaced by AI first. But understanding the business, making budget decisions and taking responsibility for results still belong to humans. That means marketers will still be needed — it’s just that marketers who use tools like AI will usually have an advantage over those who don’t.

Frequently Asked Questions about AI for Digital Marketing

How is AI used in digital marketing?

Practical uses include building easy-to-read reports, automating daily reports, analyzing which ad creatives work, finding the cause of abnormal numbers, and expanding content ideas and video scripts. Budget decisions and final checks should stay with a human.

Can AI run ads by itself?

AI agents can be set up to create campaigns and change budgets, but mistakes cost real money and AI doesn’t know the full business context. It’s safer to let AI analyze and recommend, and have a human make the decision.

Are AI-generated marketing reports accurate?

They are fast but not always correct. AI can show wrong numbers, combine results incorrectly or pull in the wrong metrics, so always check a report before sending it to a client.

What should I use AI for first in digital marketing?

Start with the most repetitive, time-consuming task, such as reporting. Once that works, gradually hand over other tasks like small automations, content ideas and analysis.

Will AI replace digital marketers?

AI will replace repetitive tasks first. Understanding the business, deciding budgets and being accountable for results still need people, so marketers who use AI well will have the advantage.

Follow Digital Break Time 
Facebook, X, Line Official Account, Instagram, Spotify, YouTube, Apple Podcast

Originally in Thai. Translated to English with the help of Claude.