Why Creating “Better” Content Doesn’t Work Anymore

Say your team spends five weeks on a definitive guide. The research is thorough, the graphics are custom and it’s longer than anything on page one. Three months after it goes live, it sits on page four and brings in about 50 visits a month.

The team followed the usual playbook. They studied the top 10 results, found the subtopics those pages missed and wrote something longer. That playbook produces a better version of an article that already exists, which gives Google little reason to rank it and readers little reason to finish it.

In this post, I’ll explain why that happens and how I plan articles now, so each one tells the reader something the ranking pages don’t.

Why “better” produces the same article

Most briefs start the same way. You search the keyword, open the top 10 results and list what each one covers. Then the team asks how to beat them.

The answers are predictable. Add more words and more examples, clean up the formatting and maybe build a calculator. Each change feels like progress, but none of them changes what the article tells the reader.

What you end up with is the 11th version of an article that already exists 10 times. It might be the best-written one. Still, a reader who has seen any of the others learns almost nothing new from it.

AI writing tools make this worse. A model writes by blending what’s already been published on a topic. Brief it with the top 10 results, and you get an even closer copy of those results, only faster.

What Google says about original content

You’ll often hear that Google “rewards information gain.” The idea comes from a Google patent called Contextual estimation of link information gain, filed in October 2018 and granted in June 2022.

The patent describes a score for how much new information a page offers compared with the pages a user has already seen. Say you’ve read two articles that cover the same ten points. A third article covering those ten points would score low. One that covers something the first two missed would score higher.

A patent shows what Google has worked on, not how ranking works today. Google’s public guidance points the same way, though. Its guide to helpful content asks whether a page provides “substantial value when compared to other pages in search results.” It also asks whether content that draws on other sources does more than copy or rewrite them.

Google has acted on this, too. After its March 2024 core update, it said search results showed 45% less low-quality, unoriginal content. A page that repeats what already ranks gives Google little reason to rank it as well.

What I found when I compared “ultimate guides”

A while back, I read several “ultimate guide” articles on B2B SaaS content marketing side by side. Backlinks and metrics didn’t matter for this. The only question was what each article said that the others didn’t.

The overlap was bigger than I expected. Close to 90% of the articles covered the same 8 to 10 core points, reordered or reworded. On average, less than 15% of each article was something you couldn’t find in the others.

Three articles stood out. Each ranked well despite having lower domain authority than the pages around it, and each added something nobody else covered.

• One shared data from a survey of 200 of its own customers.
• One went deep on a subtopic the others covered in a single paragraph.
• One challenged a common assumption and backed it up with examples and data.

None of them was “better” in the usual sense. Two were shorter than average, and one had plain design. What they had in common was that each told the reader something new.

Why teams still choose “better”

“Better” is easier to defend in a meeting. You can point to a 3,500-word guide against a competitor’s 2,800, or 12 subtopics against their nine. Stakeholders accept those numbers because they look like effort.

“Different” is a harder sell. You’re betting that the gap you found matters to readers, and you might leave out points every competitor covers.

It also takes expertise. To spot what’s missing, someone on the team has to know the subject well. That’s harder to schedule than word count, so many teams settle for coverage.

Three ways to add something the top results don’t have

When every result on page one says the same thing, I usually reach for one of these approaches.

Go one level deeper

If every ranking article is the beginner version of a topic, write the advanced version.

Say the top results all explain how to do keyword research. You could write about finding the questions keyword tools miss, like the ones people now type into ChatGPT. You serve the same interest, but you start where the reader already is.

This works well for SaaS companies with technical products. Broad “how it works” content is crowded. Content on setting up the product inside a large company with older systems is often wide open, because only people who’ve done that work can write it. The audience is smaller, but in B2B, it’s often the group closest to buying.

Correct a common claim with better evidence

Sometimes the most useful thing you can add is a well-supported correction.

Email ROI is a good example. Countless articles say email returns $36 for every $1 spent, with no context. Litmus, the source of that figure, tells a fuller story. In its 2025 State of Email report, 35% of marketers said they get $10 to $36 back, 30% said $36 to $50, and 21% didn’t measure email ROI at all. An earlier Litmus survey found that returns ranged from 32:1 for media companies to 45:1 for retail.

An article that explains where the number comes from and what a B2B team can realistically expect adds something real. If your company has its own data, you can go further. “Here’s what email returned for our customers last year, and why it’s lower than the industry figure” is something no other article can say.

This only works when you have the evidence. Experienced readers can tell when a contrarian take has nothing behind it.

Narrow the audience

Broad topics attract broad advice. When you narrow the audience, the advice has to get specific.

Instead of “content marketing for SaaS,” you might write “content marketing for API-first companies selling to developers.” That piece can name the channels developers use, the documentation they trust and the objections they raise before they’ll talk to sales. None of that fits in a general guide.

You won’t rank for the broad term. You can become the best answer for a smaller group of readers, though, and that’s often the better trade.

How I brief an article now

I still start with the top 10 results. What changed is what happens next.

1. Read the top results closely.
2. Assume the reader has already read the #1 result.
3. Write down what that reader would still need to know.
4. Note what every article skipped or rushed through.
5. Look for shared assumptions that might be wrong or out of date.
6. Build the article around those gaps.

Some people say you should write only about what’s missing. I don’t go that far. A reader who lands on your page from search may not have read anything else yet, so I keep a short section on the basics, as long as it doesn’t take over.

This takes more work up front. You need someone who knows the subject well enough to spot the gaps, which usually means bringing in a subject matter expert. That effort is also what makes the finished article hard to copy.

Original research that adds information

“Do original research” is common advice, but a lot of research adds little. A survey that asks questions with obvious answers teaches no one anything. Neither does industry data redrawn in a new chart.

Research that adds information usually does one of these things.

• It answers questions search results can’t. Plenty of reports cover what share of marketers use AI. Few show which workflows actually changed after a team adopted it.
• It tracks the same thing over time. A survey shows one moment. Data you update every quarter becomes a data set nobody else has.
• It comes from your own product or customers. Once it’s combined and anonymised, it can show how long setup really takes or where projects tend to stall.

The research doesn’t need to be large. It needs to be hard for anyone else to recreate.

What this changes for content teams

Teams built for volume tend to hire writers who are good at summarising existing sources. That produces clean articles, but rarely new information. To compete on insight, you need writers with real subject knowledge, or writers with regular access to your experts.

Measurement has to change, too. “We covered every subtopic” is a weak goal when every competitor did the same. A better test for each article is what it tells a reader that the top results don’t.

Most teams that work this way end up publishing less. Each article earns its place, though, because readers can’t get the same information anywhere else.

Ask this before you brief your next article

If someone has just read the #1 result for your keyword, what would they still need to know?

If the honest answer is “nothing, we just cover it better,” the article will struggle no matter how much work goes into it. If you can name an insight, a data point or an angle the ranking pages don’t have, the article has a reason to exist.

If you’d like help finding that angle for your own content, book a call or connect with me on LinkedIn.