What Kind Of Content Still Matters When AI Can Generate Infinite Text? (And How To Make It Move Your KPIs By 30–200%)

What Kind Of Content Still Matters When AI Can Generate Infinite Text? (And How To Make It Move Your KPIs By 30–200%)

I have spent the last 20 years helping brands turn content into revenue, not just pageviews. That includes watching three “content revolutions” up close: the rise of blogging, the social media wave, and now AI-generated text.

Each time, people said, “Content is commoditised now.” Each time, the brands that doubled down on the right kind of content pulled away from everyone else.

AI that can generate infinite text is not the death of content. It is the death of generic content.

If your articles, newsletters, or sales pages could be more or less recreated by a generic AI prompt, they will become invisible.

The good news is that your best-fit customers will still go out of their way to read, bookmark, share, and act on content that gives them something AI cannot manufacture from public data alone: your data, your judgment, your scars, and your frameworks.

Let’s talk about the specific kinds of content that will still matter, and how to design them so they move real numbers: time-on-page, scroll depth, return visitors, share rate, branded search, and revenue.

1. Proprietary Data Content That Adds 20–50% More Time-On-Page And 2–3x More Backlinks

When everyone has access to the same public information, the only way to stand out is to say, “Here’s what we see in our world that nobody else can see.”

That means original data.

Think of:

• HubSpot’s State Of Marketing report
• Zapier’s Remote Work Report
• Clearscope’s SEO ranking factor analyses
• Gong’s sales call benchmarks

These are not “10 tips for better marketing” posts. They are:

“We analysed 20,000 sales calls and found that top reps mention pricing 3 times more often than average reps, but with 25% shorter monologues.”

Numbers like that cannot be hallucinated by a generic AI. They must come from your systems, your users, your campaigns.

Why this works

When Gong started publishing data-driven content based on millions of anonymised sales calls, they saw dramatic content performance:

• Average time-on-page for their flagship data posts: 6–8 minutes, compared to a B2B blog baseline of around 2–3 minutes.
• Backlinks: individual studies earning 300+ referring domains, compared to typical B2B posts getting under 20.
• Direct leads: Gong has publicly cited that content as a major driver of pipeline, contributing to a valuation that eventually exceeded 7 billion dollars.

Readers will not ask AI, “What does Gong say about discovery calls?” They will search directly for “Gong discovery call questions” because they know the source has unique, trusted data.

How to do this in your niche

Even if you are not sitting on millions of data points, you can still create proprietary data content:

• Run small but specific studies:
– Survey 150 of your customers about their budget allocation, tool stack, or success rates
– Analyse 300 internal campaigns and compare what worked vs what failed
– Aggregate anonymised performance numbers from 50 client projects

• Turn this into content like:
– “We Analysed 247 SaaS Onboarding Flows: 3 Patterns That Increased Week-4 Activation By 31%”
– “What 187 B2B Founders Are Actually Spending On Content In 2025 (And Why 43% Plan To Cut Paid Social)”

Measure the impact

For these pieces, aim for and track:

• Time-on-page: 5–8 minutes (2–3x your baseline)
• Scroll depth: 70–80%+
• Backlinks: at least 3–5x your average
• Branded search: track increases in “brand + study name” queries over 3–6 months

When we implement this for Chedir clients, we typically set a 6–12 month goal of:

• 30–60% lift in average time-on-page for data-driven posts
• 2–4x more backlinks per flagship report than the rest of the blog
• At least 10–15% increase in branded search volume tied to the flagship series name

2. Lived-Experience Case Studies That Drive 1–3% Demo Rates And 3–5x Share Rate

Generic AI can tell a clean, theoretical story. It struggles to generate the kind of messiness and specificity that real operators recognise as authentic.

That is why detailed case studies will keep winning.

Not “Company X increased revenue by 30%.” But:

“Over 9 months, Company X increased activation from 18.4% to 27.2%, and churn from month 2–4 dropped from 7.1% to 4.3%. Here is every experiment we ran, including 3 that failed.”

Examples from real brands

Intercom’s case studies and product stories are a strong model. Their content on onboarding, in-product messaging, and support funnels routinely features:

• Before/after numbers: message response time reduced from 20 minutes to under 5; self-serve resolution up from 20% to 40%.
• Clear timeline: what they did in month 1 vs month 6.
• Real screenshots, flows, and messaging.

When firms like Intercom or Ahrefs share specific internal experiments, the performance is very different to generic “how to do marketing” content:

• Time-on-page: often 7–10 minutes for deep case studies
• Scroll depth: 70–80%+
• Share rate: 3–5x their average on LinkedIn, Twitter, or niche communities like Indie Hackers
• Conversion: it is not unusual to see 1–3% of readers of a strong, bottom-of-funnel case study click through to a product trial or demo

Why this beats infinite generic content

Ask yourself: if you run a growth team, would you rather ask a generic AI, “How do I reduce churn?” or read a piece titled:

“How We Reduced Month-2 Churn From 9.2% To 5.0% In 6 Months At A 3,400-User SaaS (Full Experiment Log Inside)”

Creators and operators know that real numbers and failed experiments are where the edge lives. AI can imitate the style, but not the substance unless you feed it that substance in the first place. Your advantage is that you own the raw material.

How to structure case studies that convert

To create case studies that your best-fit customers will spend 7–10 minutes on (and share with their peers), structure them around:

1. Clear, specific starting point
   – “Churn at 9.2% monthly, activation at 19.8%, 80% of new users never created a project.”

2. Concrete hypothesis and stakes
   – “If we could get activation to 25%, we projected an extra 32,000 in MRR within 9 months.”

3. Experiment log with numbers
   – Experiment 1: New onboarding emails → 3.1% uplift in activation
   – Experiment 2: In-app checklist → 6.4% uplift
   – Experiment 3: Removing features from the free plan → zero impact, rolled back

4. Final results and breakdown
   – “Activation from 19.8% to 27.2%, churn from 9.2% to 5.0%, net MRR +41,800 in 9 months.”

5. A clear takeaway framework your reader can steal and adapt

If you include numbers in each section, your content becomes inherently more credible than anything a generic AI can fabricate.

3. Opinionated Frameworks That Earn 25–40% Return Visitors And Above-Average Opt-Ins

Infinite AI text will push generic “how-to” advice to near-zero value. What will still attract and keep the right kind of readers is a point of view.

That means frameworks shaped by your experience, not recycled from the front page of search results.

Consider:

• Basecamp’s Shape Up method
• Drift’s early evangelism of “conversational marketing”
• ProfitWell’s approach to pricing and retention
• Ahrefs’ “no guest posts, no link schemes, just content that deserves links” stance

These companies did not just write tips. They defined a way of thinking:

“This is the way we do X. Here’s what we believe, why we believe it, and how the numbers back it up.”

Why this matters for your KPIs

Opinionated frameworks tend to behave differently in analytics:

• Return visitors: your framework series becomes something people revisit multiple times a quarter
– For strong frameworks, 30–40%+ of traffic may be returning users, compared to 10–20% on generic blogs.

• Time-on-page and scroll depth: readers know they’re consuming a “core doctrine” piece, so they stay longer and read deeper
– 6–9 minutes average time-on-page
– 70%+ scroll depth

• Opt-ins and product interest: if the framework makes sense and matches their worldview, it’s natural to subscribe or request a demo
– 3–5% newsletter opt-in on these pages
– 1–3% demo or sign-up click-through

What this looks like in practice

Imagine you run a B2B SaaS serving agencies. You could define and document a framework like:

“The 4-Layer Agency Growth Model: Lead Flow, Delivery Efficiency, Retention Levers, and Strategic Focus.”

Then create a series:

• Part 1: Why Most Agencies Stall At 1.2–1.5M ARR (And The Signals You’re Stuck)
• Part 2: The 4-Layer Model We Used To Help 11 Agencies Grow From 1M To 3–5M In 24–36 Months
• Part 3: Real Case Studies: Agency X From 1.1M To 3.4M (Retention Up From 68% To 84%)
• Part 4: Templates And Checklists: Quarterly Review Questions For Each Layer

Your goal is to make your framework the lens through which your audience sees their own problems. Then, even if they use AI, they will ask:

“Apply the 4-Layer Agency Growth Model to my situation.”

You become the mental operating system. AI becomes a tool inside it.

4. “Source Of Record” Guides That Beat Generic AI By Depth, Specificity, And Outcomes

In nearly every niche, a few sites become the place everyone points to for a certain topic:

• Ahrefs for SEO
• HubSpot for inbound marketing
• Nielsen Norman Group for UX research
• Sumo Logic or Datadog for observability best practices

These brands do not win because they can write “what is X” definitions. They win because they maintain the most complete, up-to-date, and useful resources on very specific, high-value topics.

In an AI-everywhere world, these “source of record” guides will continue to matter a lot, especially for mid- to high-stakes decisions.

Think:

• “The Complete Guide To Pricing For B2B SaaS Between 1M And 20M ARR (With 32 Real Examples And Benchmarks)”
• “The 2025 Blueprint For B2B Lifecycle Email: Benchmarks From 200M Emails And 50+ Playbooks You Can Steal”
• “The Content Operations Playbook For Teams Shipping 20–50 High-Quality Pieces Per Month”

How these guides perform

When done properly, these guides behave like:

• Pillars of your search visibility and brand
• Bookmarked resources people revisit multiple times
• Entry points into deeper relationships

Typical metrics we target with clients:

• Time-on-page: 8–12 minutes for core pillar pages
• Scroll depth: 80%+
• Return visits: 2–3x per quarter from tagged audience segments to the same guide
• Opt-ins from content upgrade or toolkit: 5–8% is realistic if the offer matches the depth of the guide
• Revenue influence: we often see 10–25% of closed-won deals touch at least one pillar guide in their journey (measured in attribution tools or CRM notes)

Why generic AI cannot replace this

AI can give a passable overview. It cannot:

• Curate only the tactics that actually worked in your specific context
• Show the nuance, exceptions, and tradeoffs you have only learned through repetition
• Keep one asset updated over years, with real user feedback from comments, sales calls, and support tickets
• Tie strategies to your product’s strengths in a credible, transparent way

These guides become more than “content.” They become living manuals for your customers’ success.

5. Newsletters As Decision Support Tools, Not “Content Feeds”

Most newsletters today are undifferentiated lists of links and ideas. AI can already replicate that.

The newsletters that will thrive are the ones that function as a decision support system for a very specific person in a very specific role.

Look at examples like:

• Lenny’s Newsletter (product managers, growth leaders)
• Morning Brew (business news with a distinct voice and curation logic)
• Stratechery (strategic tech analysis for executives and investors)

Each of these does at least one of the following:

• Interprets events through a clear strategic lens
• Synthesises scattered information into action-ready insights
• Offers frameworks you can immediately apply in a board deck or strategy doc

Lenny’s Newsletter, for example, is not just read; it is used. His longer pieces often show:

• 10–15+ minute average read times (as self-reported by readers and visible from device reading behaviour)
• Extremely high shareability in Slack channels and internal docs
• Subscription revenue strong enough to support a substantial solo business and media ecosystem

For a brand like yours

You do not have to be a media company. You can still design a newsletter that:

• Targets a specific persona (for example, “Heads of Marketing at 5–50M B2B SaaS companies”)
• Focuses on a specific outcome (for example, “de-risking your next quarter’s marketing bets”)
• Uses a consistent structure (for example, “One bet to consider, one bet to avoid, one operator story, one KPI to watch”)

When our clients treat newsletters this way, we aim for:

• Open rates: 35–50% for targeted B2B lists
• Click-through on key content or offers: 5–10%
• Content-driven sign-ups: 3–5% opt-in rates from well-matched blog posts and frameworks
• Contribution to pipeline: newsletters accounting for 10–20% of sourced or influenced opportunities within 6–12 months

Here again, the edge is your curation and judgment. AI can summarise; it cannot replace your taste and experience.

6. Sales Pages Built On Real Proof, Not Just Persuasion Formulas

AI can already write a “decent” sales page with the right prompts.

But high-converting sales pages in the age of infinite text will look different. They will be built on:

• Real proof: numeric results, before/after comparisons, cohort analyses
• Real customer language: pulled from calls, chats, surveys, and support tickets
• A clear, opinionated position on why your solution is designed the way it is

A strong sales page will feel almost like a long, well-prepared consulting session, not a string of copywriting tricks.

Consider how:

• Shopify’s Plus sales and product pages weave in real merchant stories, revenue milestones, and specific use cases
• Figma’s product/tour pages show real workflows and highlight exactly how teams collaborate, not just generic “better design faster” claims
• Notion’s pages show real templates and examples tailored to specific roles: product teams, startups, enterprise operations

Pages like these often see:

• 2–4x higher time-on-page compared to generic feature lists
• Above-average scroll depth because readers are genuinely exploring use cases and proof, not skimming headlines
• 1–3% conversion to trial or demo from cold or warm traffic, sometimes more from engaged content readers

Designing sales pages that beat AI

To build sales pages your audience will trust more than AI’s summary, bake in:

1. Data-backed claims
   – “Our customers cut time-to-value from 21 days to 9 days on average”
   – “Median increase in first-month activation: 28% across 112 onboarded customers”

2. Real-world scenarios
   – Storylines that walk through a day or week in the life of a customer before and after using your product

3. Objection handling informed by real conversations
   – Reflect back specific, nuanced worries your audience actually expresses, with evidence-backed answers

4. Connection to your frameworks and content
   – Show how the frameworks and playbooks you publish on your blog come “baked in” to your product or service

5. Making Your Website A “Source Of Record” Instead Of Just Another AI-Aggregated Site

When we design content strategies at Chedir, our goal is not to outwrite AI. It is to make your site the place people go when they want:

• The real numbers
• The real experiments
• The real stories
• The real frameworks

In other words, the material AI can only imitate, not originate.

To do that, we design content around a few key KPI targets over 6–12 months:

1. Engagement and depth
   • Average time-on-page on key pieces: 6–10+ minutes
   • 60–80%+ scroll depth on flagship articles and case studies

2. Loyalty and return visits
   • 25–40% of your content traffic coming from returning users
   • Users revisiting flagship guides or frameworks 2–3 times per quarter

3. Influence and brand demand
   • 15–30% increase in branded search queries (“your brand + topic”)
   • Content driving 2–4x more branded backlinks than baseline posts

4. Revenue impact
   • 3–5% opt-in rates from key articles to your newsletter or lead magnets
   • 1–3% conversion to trials/demos from high-intent case studies and comparison pieces
   • 10–25% of closed-won deals showing at least one touchpoint with your flagship content in CRM or attribution reports

You will not get there with volume for its own sake. You will get there by prioritising:

• Fewer but deeper pieces that could not exist without your data and experience
• Structured series of content that build frameworks, not just single posts
• Deliberate integration between your content, your product, and your sales process

Where Chedir Fits In

After two decades inside digital content, I have seen the same pattern repeat:

• The brands that treat content as a checkbox get drowned out.
• The brands that turn content into an extension of their strategy, product, and culture become the reference point everyone else cites.

At Chedir, our work is not to “fill your blog.” It is to help you create content that:

• Your best-fit customers will spend 7–10 minutes reading, not 30 seconds scanning
• They will bookmark and send to their team on Slack or email with “we need to do this”
• They will come back to multiple times a quarter when making decisions
• Will directly support higher opt-in, demo, and close rates in your pipeline

We do that by digging into your proprietary data, extracting your lived experience, uncovering niche case studies, and codifying your opinionated frameworks. Then we turn that into articles, newsletters, and sales pages that are impossible to confuse with generic AI output.

As AI fills the internet with infinite text, your audience will still be hungry for something different: sources they can trust, content that respects their intelligence, and guidance that is proven in the field.

If you are ready to build that kind of content engine, we would be glad to help you design it deliberately, measure it rigorously, and make your website the place your market keeps coming back to.