Marketers, Are You Really Ready for Generative AI?

A number caught my attention the other day, and it made me pause.

Carvana, an American used-car e-commerce company, used AI to generate 1.3 million customized videos. Every single one was different — tailored to each user's behavior as they browsed and shopped for cars.

1.3 million. Think about that. If you relied on human editors, how big a team would you need? How many all-nighters?

But AI did it. And not thrown together — each video was crafted around the user's individual browsing journey.

That got me thinking seriously about one question: How is generative AI actually changing marketing?

First, an Uncomfortable Truth

IBM ran a survey asking CMOs (Chief Marketing Officers): When do you plan to adopt generative AI?

The result? 67% said within 12 months. 86% said within 24 months.

In other words, the vast majority of marketing leaders feel this can't wait.

But here's what's interesting. Research from Deloitte shows that most companies use AI for... cost reduction and efficiency. Writing product descriptions, automating emails, trimming headcount.

Nobody's using it to innovate. Nobody's using it to grow.

It's like being handed the keys to a sports car and only using it for grocery runs.

So, How Is AI Actually Used in Marketing?

Let me walk through some concrete scenarios.

1. Chatbots and Customer Experience

What does good customer service look like?

In the past, it was "Hello, how can I help you?" — then a wait for a human agent.

Now, AI-powered virtual assistants are online 24/7, chatting with customers in natural language. A customer asks a question, they get an instant answer. Product recommendations, guided purchases — the whole journey, end to end.

Even more interesting, these AI assistants can "remember" every interaction with each customer. Someone asked about a product three months ago? When they come back, the AI knows to recommend complementary items.

Is this even customer service anymore? This is a salesperson who never sleeps, never forgets, and is online 365 days a year.

2. Content Generation

Writing a social media article — from topic to final draft — takes how long? A day? Two?

With AI: minutes. And not just text. Images, short videos, ad copy — all auto-generated.

Carvana's 1.3 million videos are proof. Adobe has also shipped ready-to-use tools like Generative Fill — type a few sentences, and your creative assets are updated.

What does this mean for marketing teams? You can produce a month's worth of creative assets in a single day, then rapidly iterate with A/B testing.

3. Personalization

User segmentation used to mean splitting people into broad buckets by age, gender, and purchase history.

With AI, it's called "micro-segmentation." How granular? Granular enough to generate personalized content for each individual user in real time.

One customer just bought a steak? AI immediately recommends a wine pairing. Another user searched for lactose intolerance? AI instantly suggests lactose-free alternatives.

We used to call it "personalization at scale." Now AI has achieved "hyper-individualization" — a thousand faces for a single person.

4. Data Analysis and Prediction

Marketing teams have no shortage of data. Social media comments, customer service chat logs, purchase behavior, browsing trails — it piles up like a mountain.

But more data doesn't mean more insight. You need someone who can spot trends and patterns in the pile.

AI excels at this. Especially with unstructured data — those chat logs, reviews, and social posts that used to give people headaches. AI makes easy work of them.

It can predict which users are most likely to convert, helping you direct resources to the right places ahead of time.

5. Workflow Automation

Scheduling social media posts, managing email sequences, translating across languages, converting file formats... AI takes over all these repetitive tasks.

Marketers can finally break free from grunt work and focus on things that actually require a human brain.

6. Creative Ideation

Kellogg's did something clever. They used AI to scan trending recipes on social media, identified ones related to breakfast cereal, and then generated fresh creative content and social posts.

In this context, AI is a creative partner that never runs out of inspiration.

Three Tiers — Where Do You Stand?

AI adoption in marketing roughly falls into three tiers.

Tier 1: Off-the-shelf tools. Open ChatGPT to draft copy, use Adobe's Generative Fill to edit an image. Low barrier, quick to adopt — accessible to individuals and small teams.

Tier 2: Custom models. Take an open-source foundation model, feed it your own brand data — historical customer interactions, product information, brand messaging — and train an AI that truly understands your business. IBM's Granite series is designed for exactly this kind of enterprise scenario, specially trained on data from legal, financial, and academic domains.

Tier 3: Enterprise-wide AI transformation. Combine multiple AI technologies to restructure entire marketing workflows. This is no longer a tool upgrade — it's an organizational transformation.

A report from the IBM Institute for Business Value (IBV) found that over half of CMOs plan to build foundation models using their own data.

In other words, many are already moving toward Tier 2 and Tier 3.

Where do you stand?

Three Hurdles

It all sounds great. But not so fast — there are three hurdles you have to clear.

Hurdle one: Data quality.

What does AI grow up eating? Data. If your data is dirty, messy, or biased, what AI produces is garbage in, garbage out. Many companies underinvest here, especially small and mid-sized businesses that lack the budget and personnel for data work.

Hurdle two: Privacy and trust.

Using customer data for personalized recommendations sounds great. But if customers feel like you're spying on them, trust collapses. Data compliance isn't optional — it's the baseline. Your AI must be transparent and explainable. Customers need to know what data you're using and how you're using it.

Hurdle three: Brand consistency.

AI-generated content gives you volume — but does every piece match your brand voice? Is the tone consistent? Is the style coherent? Many teams have stumbled here. When AI writing sounds like a "robot" instead of "you," users can tell at a glance.

One Last Thought

The opportunity generative AI presents for marketing is real. Carvana's 1.3 million videos, Spotify's cross-language podcast translation — people are already making it work.

But a tool is always just a tool. What truly determines success is the real problem you solve with AI.

67% of CMOs say they'll adopt generative AI within 12 months. But how many have figured out what they'll actually use it for?

Cost reduction and efficiency are great. But if you stop there, you're using maybe a tenth of what AI can do.

A friend of mine in consumer goods marketing said something that stuck with me: "The greatest gift AI gave us isn't saving a few people's salaries. It's that we finally dare to imagine marketing ideas we used to think were 'too labor-intensive to even consider.'"

Maybe that's what generative AI really means for marketing.