Simply adding keywords to website content is no longer enough to gain visibility. Content must now work simultaneously for people, search engines, generative AI, and AI agents.

I’ve worked in companies that published up to ten articles a week. The content often relied on the same basic facts as competitors, intending to rank for every possible keyword variation. This approach generated visibility, but rarely created deeper value.

Today, the way content is created, consumed, and optimized has changed.

AI makes information retrieval faster. The focus of website content is shifting from quantity to quality.

According to HubSpot’s 2025 research, 55% of marketers use AI in content creation, but only 7% write entire articles with it. This highlights that credibility and a human perspective are still critical.

I’ve listed concrete tips to help you:

  • increase organic visibility via generative AI
  • deliver content that empowers your brand
  • ensure chatbots can find answers on your site
  • make it easier for AI agents to act (e.g., complete purchases or fill out forms on behalf of users)

With these tips, your content will serve both people and algorithms. And your work can keep delivering results for years to come.

8 kinds of content that AI cannot create

1. Create content AI can’t easily produce

AI is good at compiling existing information, but it struggles with originality. If your content relies on the same sources as your competitors, AI won’t recognize it as valuable.

That’s why research, interviews, and experience-based analysis are so powerful. Personal stories, case studies, and practical examples help simplify complex ideas. They add unique perspectives, fresh context, and serve as evidence that something works in practice.

AI cannot generate first-hand insights about industry trends or provide local or cultural context. Internal company data, such as test results or customer feedback, becomes highly valuable when incorporated into content.

AI also cannot take value-based or ethical stances. This creates an opportunity for companies to position themselves as ethical actors through their communication.

Generative AI favors sources that bring something new to the table: original research, fresh data, in-depth interviews, or real-world experiences that can’t be found elsewhere (Andreessen Horowitz, 2025).

2. Combine data with storytelling

Numbers are just numbers until you tell the story behind them. Use data from customer feedback, user experiences, or industry trends.

For example, earlier this year, I wrote an article for Jälkisanat about the entry fee for the Finlandia literary Prize. The fee had tripled in just three years (from €100 to €300).

At first, I only had the numbers. I wanted to know why the increase happened and how publishers perceived it. That’s when the numbers turned into a story.

Open data (which is widely available in Finland) can be used to support your brand’s narrative. Even better: generate your own data and become the source yourself.

3. Provide sources and facts

Clearly present statistics, studies, and references that AI can cite. New or less frequently used sources increase the value of your content.

Be transparent about where information comes from. Citing sources boosts your site’s credibility and provides context for AI (To the Web, 2025).

In earlier SEO practices, external links were thought to “leak SEO juice.” In reality, linking out makes sense when it’s relevant to your content and your readers.

Books are powerful sources. AI has limited access to them compared to web content, and citing books doesn’t “leak” authority through links.

4. Write concisely yet with expertise

Answer the core question right away. State the main point in the first one or two sentences. Don’t bury it at the end—both humans and AI want quick answers.

Keep to one idea per sentence. This makes reading easier and helps algorithms understand your message. Each paragraph should have one clear takeaway that also works as a standalone quote.

AI values clear, fact-based writing. Aim for clarity so readers grasp your point instantly.

Pino kirjoja päällekäin

5. Use keywords naturally and semantically

Don’t repeat specific keywords mechanically. Algorithms today recognize inflections, synonyms, and related terms.

AI aims to understand the overall meaning of your content—not just exact keyword matches (IBM, 2025).

For example, if you’re writing an article optimized for the keyword “best e-bike”, you don’t need to repeat the phrase in every paragraph. Instead, you can use:

  • Synonyms (high-quality e-bike, top-tier e-bike)
  • Variations (the best e-bikes, best e-bike tested)
  • Related terms (battery, range, motor power, price-to-quality ratio)

Covering the topic broadly makes the text more natural for readers and easier for AI to interpret, as it can connect separate elements into a unified theme.

6. Use clear CTAs and buttons

Use straightforward calls to action like “Book an appointment,” “Buy now,” or “Request a quote.” These are far more effective than generic prompts such as “Read more.”

Clear language helps both users and AI agents navigate your site. Avoid vague CTAs, such as “Click here.” Better alternatives are “Explore the service” or “Book an appointment.”

Also, use HTML buttons (not images or scripts) so AI agents can reliably recognize the action.

Remember to provide internal follow-up paths after the CTA (call to action). For example, links like “Learn more about this topic” or “See case studies” guide users deeper into your site—and give AI agents multiple pathways to surface your content.

7. Show prices, terms, and delivery details transparently

Don’t hide costs—it immediately undermines credibility. Users and AI agents want to see consumer prices, delivery times, and cancellation terms upfront without unnecessary clicks or hidden sections.

Displaying prices and terms is crucial because:

  • AI agents compare services: When a user asks an AI to compare services or products, the agent only pulls from sites that clearly show pricing and terms. If those details are missing or hidden, your company is excluded from the comparison.
  • Transparency builds trust signals for algorithms: Generative AI prefers sources that look reliable and provide essential information immediately. Hidden details can suggest untrustworthiness.
  • Visibility improves user experience: When all the basics are easy to find, customers trust the site more, and AI agents can complete actions like purchases or bookings more smoothly.

Transparency costs you nothing, but it may determine whether you’re chosen or dropped from the comparison.

8. Update and enrich content

The lifecycle of content doesn’t end at publication. A website is a living whole, and algorithms evaluate it based on how well and how often it’s updated.

Generative AIs such as ChatGPT, Perplexity, and Copilot prefer fresh sources. If your content is outdated, it won’t appear in AI answers – even if it was once high-quality.

Separate annually changing information (such as pricing or regulatory updates) into distinct sections or tables to make updates easier.

Add both the original publication date and the latest update date. Updated content also signals to users that your company actively follows the industry and keeps its information accurate.

9. Use common terms

Choose language that people and AI can understand without interpretation.

Generic terms like “appointment booking” are clearer and more accessible than complicated ones (e.g., “QuickTime Booking”) or branded expressions that algorithms may not recognize.

AI agents and chatbots look for content that uses established vocabulary. If your terminology is too unusual, the agent may not understand what it means.

Defining tone of voice with 4 characteristics, which each has 7 different levels

10. Create a consistent tone of voice

A company’s voice should remain consistent across all channels – websites, newsletters, social media content, and press releases. A unified tone builds the brand and makes it easier for both readers and AI to recognize it.

Individual experts can still stand out with a personal style. They might use more casual language in a blog or on LinkedIn, as long as it fits the company’s overall communication strategy. A personal tone makes content feel more human.

Boring content doesn’t engage anyone. Make your company’s writing recognizable, and don’t be afraid to be bold. The “Did you know that…” structure is worn out. So are endless “on the one hand, on the other hand” digressions.

Think of alternatives.

What if every expert article by a CEO were written in their natural dialect?

11. Add author profiles and expert details

An author’s identity and expertise reinforce Google’s E-E-A-T principles (Experience, Expertise, Authoritativeness, Trustworthiness). An expert’s name and background information build trust and humanize the content.

Google evaluates quality partly based on whether the author is an expert in the subject. The visibility of the profile and credentials can impact how high the content ranks.

AI agents and generative search engines also prefer sources with clearly identifiable and reliable authors. Anonymous or generic content may be overlooked, since its trustworthiness can’t be assessed.

12. Use question-based headlines

Question-based headlines help tailor content precisely to user needs. Headlines like “What is…” or “How to…” directly match searches and make text easier to find for both people and AI.

If your headline is a question followed immediately by a clear answer, AI is more likely to highlight your content as a source. A question in the headline also sparks curiosity and promises readers they’ll get the answer quickly.

Use question form, especially when the topic is broad and has high search volume. Answer the question in the first few sentences, and only then expand the discussion.

Add an FAQ section to further improve discoverability and serve both people and AI. You can also create dedicated FAQ pages on key topics that focus only on the most important questions in that area.

💡 Vary your headlines: don’t structure the entire article with question subheadings, or it will end up looking like one long FAQ list.

Visualising an article or landing page content hierarchy

13. Create a clear content hierarchy

Well-structured content serves both people and algorithms. When text flows logically and follows a proper heading hierarchy (H1, H2, H3), information is easier to grasp, and AI can interpret it correctly.

A clear hierarchy can influence how your content appears in search results and featured snippets.

AI agents often look for answers to specific questions. When content is divided into logical sections, the AI can quickly find the right spot and quote it in response to a user’s query.

Practical tips for structuring text

  • Use only one H1 heading (usually the article’s main title).
  • Divide the text into clear sections with H2 headings.
  • Use H3 headings to highlight details and subtopics.
  • Build lists and tables whenever structured formats make the information easier to understand.
  • For process steps, use numbered lists instead of bullet points.

💡 Also use clear URL paths: For example, /varaa/hieronta-60min is better than a complicated address.

14. Use summaries and intros

Offer two versions of your topic: a short summary and a more in-depth version. For example, a good intro (1–3 sentences at the beginning) works as both a summary and a guide to the text, telling the reader what to expect.

Summaries and intros make content more approachable. They provide a quick overview while also promising deeper insights.

Search engines often pull answers for featured snippets directly from the intro or summary. If the beginning clearly states the main point, the page is more likely to rise to the top of search results. In addition, generative AI primarily looks for short, direct answers.

15. Summarize key points with bullet lists

Bullet lists make content easy to scan and well-structured. They help readers quickly capture the essentials – and help algorithms identify the core points.

Practical tips for bullet points

  • Keep each bullet under 50 words to ensure clarity.
  • Use strong verbs and concrete language.
  • Write lists so they also work on their own – if someone sees only the list, they should still grasp the main idea.
  • Add references at the end of the list when needed to strengthen credibility.

Remember to include an introduction to the lists and elaborate on the subject later in the text. If you overly emphasize lists, you fail to give a thorough inspection of the topic.

Travel insurance cluster showing interconnections between themes

16. Build content into clusters

Link related pages together thematically and with internal links. When your site forms a clear cluster – a content ecosystem – it works like a spider’s web: each piece of content supports and strengthens the others.

AI agents and generative search engines don’t just look for individual articles but for entire ecosystems. If your content forms a logical network, AI can use it more effectively as a source.

17. Make content accessible

Good accessibility benefits both people and AI. When content is clearly marked and easy to navigate, it reaches more users – and is better understood by AI.

Practical steps for web accessibility

  • Metadata: Keep meta titles within 50–60 characters and meta descriptions within 150–160 characters. This improves search visibility and communicates the core idea to both users and algorithms.
  • Alt text: Add descriptive alt text to images. It helps screen readers, improves accessibility, and explains what the image shows.
  • Clear language: Avoid overly complex terminology and write so the message is immediately clear.
  • Color contrast and visual readability: Follow WCAG (Web Content Accessibility Guidelines) to ensure readability for people with visual impairments and across different devices.

Generative AI interprets content more effectively when it has solid metadata and a clear structure. For example, alt text can help AI understand what an image shows and surface it in a response.

18. Offer content in multiple formats

Use a variety of content types: text, tables, images, video, and audio. This way, you serve different types of users – visual learners, listeners, and readers – while also improving your visibility in AI-driven searches.

AI doesn’t necessarily extract an entire article but may pull specific elements. A table or infographic, for instance, can work as a standalone answer in an AI search.

19. Use visual elements

Visual elements such as infographics, charts, and illustrations make content easier to grasp and remember. They support the text and condense complex ideas into clear, visual form.

Use infographics to illustrate processes or steps. Add charts that compare options, show trends, or summarize data.

Incorporate your brand’s visual identity so that images reinforce a consistent tone of voice.

A table comparing reactive and proactive content production strategies

20. Shift your content strategy from reactive to proactive

A reactive content strategy responds to what users are already searching for. A proactive content strategy looks ahead: it provides answers before the question even arises.

Generative AI favors content that is timely and diverse. If your site covers topics others haven’t yet addressed, AI is more likely to cite you as a source.

Search engines also reward sites that are among the first to cover emerging and trending topics. This gives you a competitive edge and helps establish authority in your field.

Follow industry trends, research, and forecasts. Turn them into clear articles, blogs, or guides. Publish content that helps customers prepare for changes – e.g., upcoming regulations, new technologies, or shifts in customer behavior.

Fewer publications, more depth and value. That’s how your content remains relevant.

Keyword stuffing and superficial texts are no longer effective. Instead, long-lasting evergreen content linked to broader content clusters can continuously attract traffic and build search engine trust.

Generative AI cites sources that are detailed and include references, data, and examples. If your content doesn’t add anything new, AI won’t mention it – and you’ll lose visibility.

💡 For a company, it can be more effective to publish one comprehensive, well-sourced guide (e.g., “How to choose the right CRM system?”) than five shallow blog posts on the same topic. Search engines and AI recognize that the guide is a thorough source and are more likely to highlight it.

21. Ensure fast loading and mobile-friendliness

Speed and mobile usability are critical for both user experience and AI optimization. Slow-loading pages drive users away and push your site down in search rankings.

AI agents prefer sites where information is quickly accessible and easy to read on mobile. If your content hides behind heavy scripts or large files, AI may skip it.

Practical tips

  • Optimize images: compress them without losing quality.
  • Test page speed with tools like Google PageSpeed Insights or GTmetrix.
  • Ensure responsiveness across devices and screen sizes.
  • Minimize unnecessary scripts and heavy elements that slow down loading.

22. Use schema markup

Schema markup is additional code (structured data) you can add to your website’s HTML. It tells search engines and AI systems what your content means – not just what it says.

Think of schema as a dictionary that helps machines interpret content the way humans do. There are many types of schema depending on the use case.

Examples of schema types:

  • Product → price, stock availability
  • How-to → step-by-step instructions
  • FAQ → question + answer pairs
  • Service → duration, location
  • Event → date, venue, tickets
  • Reservation/Action → bookings and purchases
  • LocalBusiness → local visibility

For example, Booking.com uses schema markup on hotel pages to display: price, location, reviews, and availability. That’s why their listings appear in Google results with direct price details.

Not all schema types are worth adding. The right ones depend on your industry and goals.

23. Support universal payment methods

Google Pay, Apple Pay, PayPal, and credit cards simplify the buying process for both users and AI agents.

Offering multiple, universal payment methods makes the checkout process smoother. The less friction at checkout, the more likely the purchase.

Display payment options clearly on the product page or before checkout. Provide at least one mobile payment option (e.g., Apple Pay or Google Pay).

Support the most popular international payment methods if you sell abroad.

Visualising how 1 to 2 click checkout process looks like in practice

24. Shorten the purchase journey

Aim for conversion in just 1–2 clicks: from the product page straight to cart and payment.

Eliminate unnecessary steps such as mandatory registration, lengthy forms, or repeated confirmations. Offer guest checkout so that a user or AI agent can purchase without creating an account. Provide “buy now” buttons directly on product pages.

The shorter and clearer the purchase process, the more likely customers are to complete it. The same applies to AI agents: if the path to purchase is clear, they can navigate it all the way through on behalf of the user.

You can test the purchase journey yourself regularly to see if it’s frustrating for your customers as well.

25. Make forms simple

Ask only for essential information: name, email, address, and payment method.

Use clear field labels, not vague abbreviations or internal jargon. Also take advantage of autofill features (e.g., browser or mobile suggestions).

A clear, quick form makes transactions smoother for both people and AI agents. The fewer unnecessary fields, the more likely the user is to complete the process.

Before creating new content: optimize the current

HubSpot analyzes and updates its existing blog articles annually. According to their 2021 statistics, about 76% of their traffic came from updated, older content – not from new articles (HubSpot 2021).

Most websites have dozens or even hundreds of articles, but only a fraction generate traffic. Before writing new texts, analyze the following:

  • Which content currently drives traffic (Google Search Console / GA4)?
  • Which keywords are essential for your business? Which content already ranks high in SEO (Google Search Console / SEO tools)?
  • Which content is already appearing in AI search results (Perplexity, ChatGPT, Copilot)?
  • Which articles contain outdated information, broken links, or weak CTAs?
  • Should some pieces of content be merged?
  • Which important topics and themes are missing from your content?
  • Are internal links between articles missing?
  • When was the last time you enriched existing content with charts, statistics, examples, or other information?

Go through these points before creating new content, as they can deliver quick wins. AI favors fresh and relevant content, and algorithms rank your site as more trustworthy.

Note: Organic traffic is currently decreasing due to Google’s AI summaries that show the answer already on the front page of the search results. Therefore, the user doesn’t have to click on the website.

Hold regular content workshops

Every content team should hold 2–4 content workshops per year. The purpose of a workshop is to review the current state of your content, decide on upcoming directions, and enrich existing material together.

Before the workshop, consult company leadership about business goals and strategy: what are the key priorities the company is focusing on? Website content should reflect those priorities.

Workshops help deliver concrete results and keep content creation purposeful. In many companies, planning content falls on just 1–3 people. Staying inspired and creative year after year with such a small group is a challenge.

Need help with content optimization?

Do your current website texts need fine-tuning? Or do you need support in creating a new content strategy to better align with your business goals?

I can help – from strategy development to editing and producing content. I also run content workshops.