AI is advancing so quickly that it’s reshaping marketing in real time. In just a year, it has evolved from a simple curiosity into a powerful tool that cuts hours of work, boosts efficiency, and enables smarter strategic choices.

Even basic AI tools can now handle 50–70% of a marketer’s daily routine — from selecting targeting options and analyzing ad campaigns to working with educational materials, technical documentation, or lists of placement channels.

In this article, we’ve gathered real use cases from the newage. team. Everything here is based on practical workflows you can apply immediately after reading.

AI in Audience Building

One of the strongest basic applications of AI is constructing and refining target audiences.

Normally, creating a customer persona takes a lot of time: gathering product context, forming hypotheses, analyzing behavioral patterns, and comparing competitors. Artificial intelligence reduces this process several times while increasing the quality and depth of insights.

This material is based on a webinar hosted by Sasha, the CEO of newage. You can watch the full recording below:

In our work, we use our own Prompt Base: the marketer provides just a few key parameters — the product, category, campaign goals, market, and competitors. The model then independently assembles all the inputs into a complete audience profile. AI immediately builds segmentation, describes motivations and barriers, highlights behavioral patterns, and explains decision-making context.

What AI generates automatically:

• demographic characteristics
• psychographics and motivations
• pains, triggers, barriers
• behavioral patterns across the funnel
• interests, context, and reasons why this audience buys

A particularly strong capability is how AI breaks down the audience using the SEE–THINK–DO–CARE model:

• SEE: broad audiences who don’t yet know the product
• THINK: people with an emerging need
• DO: hot segments ready to act or purchase
• CARE: existing customers who require retention

He doesn’t just generate surface-level “cold/warm” segments — he explains what motivates a person at each stage, what stops them, how their intent evolves, and which formats or messages will work best.

For example, in THINK, artificial intelligence can suggest messages that strengthen comparisons; in DO, arguments that dispel fears; and in CARE, ways to increase loyalty.

AI works exceptionally well with details. It can highlight specific emotional drivers, identify realistic barriers, suggest interconnected triggers, and immediately recommend the right content format for each case.

When working with the same product for a long time, a marketer’s vision inevitably becomes “blurred.” We tend to stick to audiences that have worked before and overlook new opportunities.

AI, on the other hand, suggests segments that go beyond the obvious: adjacent categories, alternative behavior scenarios, and interests that have never been considered.

Very often, these new groups become the source of growth in campaigns — and this is one of the most valuable effects of using AI. The combination of speed and depth gives marketers a strong advantage: audience portraits become more accurate, segmentation becomes broader, and strategy becomes more grounded. AI removes routine work and creates a high-quality foundation for communication and media planning.

AI in Audience Selection and Keyword Discovery

Once we understand who our audience is and where they are in the funnel, the hardest part begins — figuring out how to actually reach them.

Google Ads, Meta, and TikTok offer tens of thousands of targeting options, interests, search queries, and behavioral signals. It’s hard for a person to cover this entire volume, while AI handles it incredibly easily.

AI takes over the routine of search: simply describe the product, campaign objective, and funnel stage, and the model returns a curated set of targetings to test.

Important: AI does not invent targeting; it finds it in pre-uploaded Google Audiences or Meta Interests files provided as input data.

In addition, AI excels not only at selection but also at cleaning keywords. A marketer can upload a large array of search queries, and AI:

• identifies irrelevant options,
• explains why they are not suitable,
• proposes alternatives,
• structures the campaign by user intent.

A human reads the keywords literally, while AI sees the logic, similarities, and broader semantic connections. For example, for the category “renting housing,” AI could suggest segments related to moving, renovation, insurance, and furniture searching.

How AI Saves Tens of Hours and Strengthens Brand Safety

Building high-quality whitelists and blacklists takes years of work. In newage., this database has been developed over the past 7–8 years and includes over 30,000 YouTube channels. Most specialists of such resources simply do not have them, but now AI can compensate for that.

To manually curate a list of channels for placements, a marketer would need to:

1. find thematically relevant channels;
2. review their content;
3. assess audience quality;
4. verify brand safety;
5. filter out toxic or dubious sources;
6. compile the final whitelist or blacklist.

This is a monotonous process that easily consumes days or weeks. And the worst part is that it quickly becomes outdated: content changes, new channels appear, old ones stop publishing.

AI takes on the part of the work that drains the most energy — sifting through large arrays of channels and the initial level of filtering. The algorithm looks something like this:

1. The marketer describes the task: “need business channels for real estate promotion,” “need channels about investments,” “need thematic TikTok profiles.”
2. AI analyzes available channel data: descriptions, topics, keywords, and content character.
3. It forms lists that match the themes and audience level.
4. It filters out anything potentially toxic, dubious, or irrelevant.

Ethical Aspect: Avoiding Russian Content

A separate topic worth noting: cannot sponsor Russian content, even accidentally. Why?

• Displaying ads monetizes authors → they pay taxes → the funds go, you know where.
• This creates a false impression of a “large Ukrainian audience” on Russian channels.
• This directly harms brand reputation.

AI can help automatically filter out:

• content in Russian
• topics related to the Russian media sphere
• suspicious channels with mixed audiences

For example, AI can identify channels with Russian-language descriptions — a quick way to automatically exclude a large portion of unwanted content. As a result, brands not only appear where “possible” but do so consciously, responsibly, and effectively.

AI-powered Campaign Analysis: a New Level of Speed and Insights

After launching an ad campaign, the next step is analysis of results. In theory, this sounds simple, but in practice, it often boils down to scanning dozens of lines in Google Sheets, looking for patterns, comparing ad groups, CTR, CPA, ROAS, audiences, devices, formats, and a dozen other metrics. This is exactly where AI delivers the most noticeable gains in speed and quality.

How AI Works with Google Sheets Data

After the important data anonymization step, the sheet is uploaded to AI. The model recognizes the structure: campaign names, ad groups, key metrics, audiences, devices, formats, and dates.

In mere seconds it:

• identifies the strongest and weakest campaign elements;
• compares them against each other;
• forms initial conclusions;
• suggests what to verify or optimize.

This results in not just an interpretation of numbers, but a concise and structured snapshot of performance.

People analyze data from the perspective of experience, product, market, and context. They understand why a certain change occurred and sense the user’s logic. AI, in contrast, has no biases, is not overwhelmed with information, and is not tied to past decisions.

Another interesting feature: AI not only analyzes data but also suggests what to do next. Based on the data, the model can propose several optimization directions:

• which audiences are worth testing again;
• which creatives are “dropping” and need replacement;
• where it makes sense to change bidding type;
• how to reallocate the budget between groups;
• which segments to disable due to low effectiveness.

AI does not replace analysis but provides inspiration and new angles often overlooked by habitual viewpoints. Analysis stops being painful and becomes a regular, structured process.

Generation of Formulas, Scripts, and Automations

If a digital marketer’s work is described honestly, a significant part of it is not creativity or even strategy. It is routine: preparing spreadsheets, checking data, filtering lists, creating formulas, and spotting errors in reports. AI has become a tool that literally “frees” the marketer from this routine.

Complex Google Sheets Formulas in One Request

When a table grows to hundreds or thousands of rows, formulas become a major headache. VLOOKUP, QUERY, REGEX, SPLIT, IMPORTRANGE — all work well, but often require lengthy testing and debugging.

AI allows:

• to describe in words what the formula should do;
• to obtain a ready-made variant;
• to immediately see an explanation of the logic;
• to adapt it to the user’s own spreadsheet structure.

Simple, but very illustrative example: it’s needed to filter YouTube channels whose names contain the Russian letter to remove such content automatically. The script did this in seconds.

Fewer Mistakes, Less Routine Work

When formulas are created manually, errors are the norm: an extra parenthesis, a missing comma, an incorrect range. AI effectively eliminates this level of error. The marketer stops spending an hour searching for a bug that would be spotted in seconds by AI.

All that is required from a specialist is to set the task precisely. This becomes a new key skill: not programming, but framing the commands. Because the result depends on how well the AI is guided about the desired output.

Accelerating Learning

In digital marketing, learning never stops: new trends, updates to tools, case studies, webinars, lectures, and podcasts. The problem is just one: there isn’t enough time for all of it. AI finally makes learning manageable and takes on the most burdensome part — reviewing and summarizing.

Uploading a Lecture or Webinar and Getting a Summary

Everything works as simply as possible: upload a video, audio, or just a link, and AI transcribes the material into text, structures it, creates a summary, and generates a list of key ideas. In other words, instead of a two-hour webinar, you get the five-minute essence.

This is especially useful for long Google or Meta trainings that hardly anyone has time to attend.

After the summary, it’s possible to work with the lecture the same way as with a live expert. AI allows asking clarifying questions. Essentially, this creates a hybrid — a lecture plus a personal mentor who explains the material in your context.

And what about video? AI doesn’t just retell them. It:

• highlights the main theses,
• filters out the “fluff”,
• keeps only what can be applied in work,
• suggests which insights to implement in campaigns.

AI can independently identify points relevant for competitive analysis, YouTube channels, and audiences. Another highly practical aspect: AI can assess whether it’s worth spending time on a full review at all.

That is:

• if there is something new or unique in the material, AI highlights this;
• if it’s “a repeat of previous training sessions,” it honestly states that a summary is enough.

On average, time savings reach 60–70%, and sometimes even more. Learning stops being a “burden with no time” and becomes part of the workflow that AI takes on.

More about working with AI:

• AI Arsenal of a Marketer: from Creatives to Campaign Automation
• How to set up the AI workflow process

Information Search in Google Ads, Meta, and DV360 Help Centers via AI

Searching manually is a pain. That’s why today the simplest solution is AI, which literally finds exact, verified, and up-to-date information straight from official sources in seconds.

For AI to provide a correct answer, it’s important to formulate the query clearly. What do you want to learn? In what context? What additional criterion matters? The more precise the question, the higher the quality of the result.

The advantage of information search in help centers through AI is that it does not provide an answer. It works with what is provided, and the process looks like this:

• you paste the URL of the official help center,
• formulate the query,
• AI reviews the documentation,
• finds the exact fragment,
• and returns a concise answer plus quotes from the sources.

For a marketer, this means saving up to 90% of the time usually spent on technical searches. Instead of long navigation between sections, scrolling, and comparisons, you simply receive a ready-made answer that already captures the essence.

Modern AI on tools that eliminate routine, accelerate analysis, help think more broadly, and turn data chaos into structured decisions. Even basic use cases deliver a time saving of 50 to 90%.

In the next materials, we will move on to advanced AI tools that the newage team uses in complex advertising projects.

FAQ: Common Questions about AI in Marketing

Can AI completely replace a marketer?

No. AI handles routine tasks well: segmentation, filtering, analysis, and formulas. But strategic decisions, data interpretation, market understanding, and brand context remain human responsibilities.

How accurate are AI recommendations?

At a basic level, very accurate. But any result needs to be interpreted and verified.

Is it safe to upload advertising campaign data?

Yes, if anonymization is performed in advance: remove client names, brands, and sensitive information. This is mandatory.

Can one prompt be used for all projects?

Yes and no. There are ready-made templates, but they must be adapted to the product, category, market, and business goals. The same prompt without context will yield mediocre results.

Does a marketer need to know how to code to automate work?

No. In 2025, AI takes care of writing formulas and scripts. The only real skill is knowing how to formulate the task clearly: what needs to be obtained and why. AI will handle the rest.