The future of AI and its intersection with marketing

My career in content and SEO started back when SEO meant something completely different. SEO best practices used to mean:

• Use the target keyphrase about 15-25 times in a standard-length article
• Ensure 60-75% of headings include a version of the target keyphrase
• Use related terms to the target keyphrase in headings and paragraph copy as much as possible
• Make sure image alt text pertains to the target keyphrase
• Include similar headings and concepts to the top five articles that rank for this search term
• Get backlinks from authoritative sources that link to your content with the appropriate anchor text
• Get internal links from your own authoritative pages
• Don’t link to competitors or other pages that rank for similar keywords

These were flawed. I see a lot of SEOs on LinkedIn that reminisce about the “good old days” before AI. The era those people are referring to was artificial, gameable, imperfect and didn’t deliver a good end-user experience. SEO sucked back then.

Some of the inherent downsides to SEO articles back then included:

• It’s fine to have walls of text that are keyword optimized
• Say all the right words even if they’re shallow: Define them but don’t waste page space elaborating
• Focus on talking about related keywords rather than things the user might be interested in
• And so on

Ultimately, the goal of SEO content was to game search engines, not provide value to Google. The way this resulted in revenue was often that the user didn’t understand enough from the article alone, and had to click through to your “good pages” to learn more.

Think about that user experience for a second — having to sniff out the meaningful web pages like a truffle pig.

Aside: I used to order pizza from a restaurant whose order page had a paragraph at the bottom that said “Pizza near me, food delivery near me, pizza delivery Ann Arbor, Michigan, pepperoni pizza, pepperoni pizza near me, customizable pizza near me, order pizza near me, pizza delivery online, online pizza ordering…”

Search engines are smarter now

Before forming an opinion about how good or bad AI is, it’s important to understand that search engines themselves have evolved a lot since then.

Almost all of Google’s updates over the last decade have been called some form of “Helpful Content Update,” which gives a good indication of what they’re expecting, indexing, and thus, what behaviors you should avoid.

These updates have slowly but surely chipped away at the traditional SEO techniques that marketers used to employ to game the system.

Have you ever Googled something for a quick answer but not found it? Instead of finding the simple answer you’re looking for, you find a long article that talks about every related bit of information to the answer and defines all the core concepts related to the answer you’re looking for. Recipes are notoriously guilty of this.

What “helpful content” means

In plain English: Helpful content is content that does the thing it promises to do — quickly, clearly and without the runaround. No secret handshake, no magic formula, no keyword bingo. Helpful content gives users what they came for and doesn’t make them hunt for it in a haystack of SEO tricks or word count padding.

“Helpful” is less about how you write, and more about who you’re writing for. It’s about serving the reader, not the algorithm.

Here’s the gut-check: if someone lands on your page, can they leave genuinely satisfied? Did you answer their question, solve their problem, or make their life easier — even if only a little bit? If yes, congrats: you’re making helpful content. If not, you’re probably making content for bots and praying it sticks.

Google’s definition of helpful content

“People-first content is content that’s created primarily for people, and not to manipulate search rankings. It answers real questions, demonstrates expertise, and leaves the reader feeling they’ve had a satisfying experience.”

According to Google, helpful content is:

Created for people, not search engines: Content should be written to genuinely help real users, not to game rankings.

Meeting the needs of the visitor: Answers the user’s query fully and clearly, leaving them satisfied (not needing to look elsewhere).

Demonstrating first-hand expertise: Shows real experience, authority, or a depth of knowledge about the topic.

Original and substantial: Provides unique insights, information, or value not found elsewhere — avoid just repeating what’s already out there.

Trustworthy and accurate: Information should be reliable, well-sourced and factually correct.

Easy to read and navigate: Organized and clear, with minimal distractions or excessive ads.

AI tools are also smarter now

Search engines aren’t the only things that have evolved — AI content tools are on a completely different level than the “article spinners” or keyword stuffers of the past.

“Write me a blog post about X” is honestly the least interesting thing you can do with them now.

ChatGPT is basically a pocket-sized strategy partner. Beyond drafting copy, it can:

• Analyze a dataset and summarize your key takeaways in seconds
• Roleplay as a skeptical customer and stress-test your messaging before you go live
• Instantly turn dense research into an executive summary or email
• Compare large SEO tables exported from other tools and devise a clear strategy

Notion AI does a lot more than auto-generate meeting notes or task lists. You can use it to:

• Summarize sprawling project documentation for new teammates, instantly
• Turn unstructured SEO strategies into actionable project plans or checklists

AirOps (and similar workflow automation platforms) take things even further by connecting your AI directly to your data and processes. For example, you can:

• Auto-generate personalized email sequences for every segment in your CRM, triggered by live data
• Extract, transform and analyze large volumes of data (from spreadsheets, databases or APIs), then summarize or visualize the results
• Sync real-time insights from across marketing, sales and support to build dashboards or trigger alerts — without manual wrangling
• Write blog posts that are trained on your style guides, your product, your knowledge bases, and instantly publish them through the CMS

The main difference with AI now is that it isn’t just a lukewarm content vending machine anymore — it’s a co-pilot for thinking, creating and executing across every layer of your workflow.

Where marketers miss the mark

I’m so tired of seeing people saying “How is AI gonna replace me? It can’t even count the Ns in the word ‘banana’!”

First of all, cringe.

Secondly, this is a false conclusion, and is brought about by asking the wrong question — yet these posts get so much traction on LinkedIn, which is discouraging because it shows that people still aren’t using AI the right way.

Unless you’re a really shallow writer, you shouldn’t worry about AI replacing you anyway.

AI helps you strategize, plan, measure and produce more content than ever before.

What’s more is that, with tools like AirOps or Scout and Contentstack, you can configure workflows that integrate with your CMS, meaning:

• Automatically create and update CMS entries like blog posts, landing pages, product listings and more
• Generate Schema for each CMS entry that conveys key information directly to search engines
• Pull and analyze reporting and automatically plan what content to create next

Basically, AI is like a team of assistants, and you’ve been promoted to a managerial position. (Congrats btw)

What will the future look like?

Everyone becomes an AI user

In the near future, everyone working in content or marketing will need to understand at least the basics of AI.

Instead of assembling a traditional team of writers, it will often just be you and a set of AI tools working in tandem. This shift may feel strange at first, but it’s essential to learn to work with AI as your collaborator rather than resist the change.

As AI-assisted search becomes the norm, we’ll also see large language models (LLMs) take over many question-answering tasks that used to belong to search engines. In fact, ChatGPT just released a Chrome extension that lets users replace Google with it.

Helpful content still matters

Despite all the advances in automation and AI, helpful content will remain king. High-quality, valuable content is what gets included in LLMs; poor content is simply ignored or filtered out.

This means your content probably won’t show up in the way you intended. For example, instead of someone reading your full blog post, they might find your insights embedded in a step-by-step guide that ChatGPT generates for them — and that’s okay.

The key is that your content needs to be genuinely useful, no matter how it’s discovered or delivered. Unhelpful content, on the other hand, will increasingly serve no real purpose and will quietly disappear from the digital landscape.

An internet saturated with AI

We’re heading toward a reality where the vast majority of blog posts and similar web content will be at least partially generated by AI.

In practice, this could mean marketers and content creators draft a barebones outline and let AI fill in the dates, details or routine information, streamlining the whole process. For many companies, 100% AI-generated SEO content is already the standard.

Human writers will remain in the loop, but their role will often shift to overseeing, editing or guiding AI-generated drafts rather than writing everything from scratch.

Author’s note: AI automation tools like Scout and AirOps are by far the most important tools I’ve ever learned to use in my marketing career.

Human connection becomes more important

As I predicted in Contentstack’s Digital 2030 report, AI-generated content will become ubiquitous, with readers realizing that most blog posts are written by a machine.

Because of this, social proof — testimonials, case studies, real user stories — will become even more valuable. People will naturally gravitate toward personalities, influencers and creators who bring a unique voice, strong opinions or an authentic perspective to the table.

While AI handles the creation of neutral, informational content, it will be up to humans to provide personal stories, creative takes and content that entertains or connects on a deeper level.

The future will belong to those who can combine the efficiency of AI with the irreplaceable spark of human insight and connection.

Jack Virag leads Content at Tavus, a research lab on a mission to make technology feel human, with video AI that can see, hear and converse in real time. Follow him on LinkedIn.