Most SEO teams still rely on checklists and guesswork. Learn how a hypothesis-driven approach leads to faster and smarter results.Most SEO teams still rely on checklists and guesswork. Learn how a hypothesis-driven approach leads to faster and smarter results.When it first became clear that clicks and user behavior influence rankings, many in the SEO industry dismissed it – until it became widely accepted.
Today, the SEO vs. GEO debate is sparking similar reactions.
It’s a healthy conversation, but it also highlights gaps in how we think as a community.
We don’t lack checklists. What we lack is an open mindset – a willingness to admit we don’t know everything about search or LLMs.
Acknowledging that is the first step. From there, we can adopt the mindset we need: one rooted in testing.
Marketers underuse experiments
Experiments are one of the most effective ways to uncover what actually works – yet most marketers don’t use them.
In a Harvard Business Review article, the author notes that only a third of senior digital growth managers had ever run digital ad experiments.
While the article focuses on advertising, the same applies to SEO. It’s still rare to see SEOs build testing into their strategy.
In my experience, even explaining what an SEO test is – to clients or in-house teams – can be a challenge.
There’s a common misconception that a test is only successful if it “passes,” when in reality, both passing and failing yield valuable insights.
Unfortunately, a failed SEO test is often mistaken for failed SEO work.
Then there’s the developer side – full sprints, backlogs, maintenance, technical debt.
Asking for extra work to test something uncertain isn’t exactly a popular request.
All of this contributes to the hesitation – and in some cases, resistance – around testing in SEO.
Dig deeper: SEO testing: Shifting from reactive to proactive strategies
How to perform an SEO test?
There are plenty of experimentation frameworks from product and CRO that can be adapted for SEO.
But for now, the focus is on a data-driven method: the hypothesis-driven approach.
What is a hypothesis?
A hypothesis is an assumption you can test.
It’s a clear, specific statement about what you expect to happen and why.
Most hypotheses follow an if-then format. For example:
“If we fix broken redirects, then organic traffic will improve.”
Here’s a simple way to illustrate the process:
The problem: I don’t know which SEO course to take.
Hypothesis: If I ask in three different Slack groups, then at least one will offer a good recommendation.
You test the hypothesis by asking in those Slack groups.
If you get a helpful answer, you accept the hypothesis. If not, you reject it and come up with a new one.
It might seem simple, but this is a structured, statistical way to test assumptions – and one we should use far more often in SEO.
Dig deeper: 6 SEO tests to help improve traffic, engagement, and conversions
Applying a hypothesis-driven approach to SEO
Let’s say a client hires you because they lost traffic after a migration and need help recovering it.
A traditional SEO might start with audits and checklists – a “pray and spray” approach I often talk about. You try to do everything right, hoping something works and traffic returns.
But this method is time-consuming, costly to execute, and offers little learning.
You’re applying the same checklist to every site, without considering new tactics or gaps in your current workflows.
The hypothesis-driven approach, by contrast, is more strategic and data-informed.
You begin by listing all potential causes based on your experience, analysis, and knowledge:
Broken redirects.
Content changed by more than 30%.
JavaScript issues.
Page speed issues.
Then, you choose one as your starting hypothesis – the issue you believe is most likely. For example:
“If we make sure all redirects are working, then the traffic will recover.”
You audit the redirects, identify any that are broken or missing, and recommend fixes.
If traffic begins to recover, your hypothesis is valid, and the dev team spent minimal effort on it.
If not, you move on to the next most likely issue.
Advantages of a hypothesis-driven approach to SEO
I’ve always been an advocate of thinking strategically and using data – and this approach supports both.
Here’s why starting with a hypothesis and testing it is, in my view, the best way to approach SEO:
It keeps the scope manageable: You start with limited knowledge, so you’re not diving into every possible SEO issue. The search for answers becomes more focused and efficient.
It avoids analysis paralysis: While you’re still using data (e.g., checking for broken links), you’re narrowing your attention to a few likely causes rather than trying to fix everything at once.
It’s iterative: If the hypothesis doesn’t hold, you refine your thinking – drawing from case studies or others’ experiences to shape the next test – rather than blindly running through a checklist.
It leads to real learning: This approach helps uncover meaningful insights, like: “If your Core Web Vitals scores dropped significantly during a migration, then fixing them can help recover traffic.”
Dig deeper: 5 practical SEO experiments with AI as a co-pilot
Progress requires a testing mindset
Adopting a testing mindset doesn’t just improve existing strategies – it opens the door to discovering new ones.
Take the example of a Bing employee who had an idea to change how ad headlines were displayed.
The idea was buried in a backlog and marked low priority – until an engineer ran an A/B test and uncovered a 12% lift in revenue. Without that test, the idea would’ve been lost.
Despite how quickly the SEO landscape is evolving, there’s still surprising resistance to new ideas.
It often seems that there’s only one way to do SEO, and it starts and ends with checklists.
Whatever your stance, the real question is: Have you tested it?
You don’t need to adopt every new concept, but you do need to evaluate them with an open mind.
Let’s not default to skepticism just because something’s unfamiliar. Let’s test more.
When it first became clear that clicks and user behavior influence rankings, many in the SEO industry dismissed it – until it became widely accepted.
Today, the SEO vs. GEO debate is sparking similar reactions.
It’s a healthy conversation, but it also highlights gaps in how we think as a community.
We don’t lack checklists. What we lack is an open mindset – a willingness to admit we don’t know everything about search or LLMs.
Acknowledging that is the first step. From there, we can adopt the mindset we need: one rooted in testing.
Marketers underuse experiments
Experiments are one of the most effective ways to uncover what actually works – yet most marketers don’t use them.
In a Harvard Business Review article, the author notes that only a third of senior digital growth managers had ever run digital ad experiments.
While the article focuses on advertising, the same applies to SEO. It’s still rare to see SEOs build testing into their strategy.
In my experience, even explaining what an SEO test is – to clients or in-house teams – can be a challenge.
There’s a common misconception that a test is only successful if it “passes,” when in reality, both passing and failing yield valuable insights.
Unfortunately, a failed SEO test is often mistaken for failed SEO work.
Then there’s the developer side – full sprints, backlogs, maintenance, technical debt.
Asking for extra work to test something uncertain isn’t exactly a popular request.
All of this contributes to the hesitation – and in some cases, resistance – around testing in SEO.
Dig deeper: SEO testing: Shifting from reactive to proactive strategies
There are plenty of experimentation frameworks from product and CRO that can be adapted for SEO.
But for now, the focus is on a data-driven method: the hypothesis-driven approach.
What is a hypothesis?
A hypothesis is an assumption you can test.
It’s a clear, specific statement about what you expect to happen and why.
Most hypotheses follow an if-then format. For example:
- “If we fix broken redirects, then organic traffic will improve.”
Here’s a simple way to illustrate the process:
- The problem: I don’t know which SEO course to take.
- Hypothesis: If I ask in three different Slack groups, then at least one will offer a good recommendation.
You test the hypothesis by asking in those Slack groups.
If you get a helpful answer, you accept the hypothesis. If not, you reject it and come up with a new one.
It might seem simple, but this is a structured, statistical way to test assumptions – and one we should use far more often in SEO.
Dig deeper: 6 SEO tests to help improve traffic, engagement, and conversions
Applying a hypothesis-driven approach to SEO
Let’s say a client hires you because they lost traffic after a migration and need help recovering it.
A traditional SEO might start with audits and checklists – a “pray and spray” approach I often talk about. You try to do everything right, hoping something works and traffic returns.
But this method is time-consuming, costly to execute, and offers little learning.
You’re applying the same checklist to every site, without considering new tactics or gaps in your current workflows.
The hypothesis-driven approach, by contrast, is more strategic and data-informed.
You begin by listing all potential causes based on your experience, analysis, and knowledge:
- Broken redirects.
- Content changed by more than 30%.
- JavaScript issues.
- Page speed issues.
Then, you choose one as your starting hypothesis – the issue you believe is most likely. For example:
- “If we make sure all redirects are working, then the traffic will recover.”
You audit the redirects, identify any that are broken or missing, and recommend fixes.
If traffic begins to recover, your hypothesis is valid, and the dev team spent minimal effort on it.
If not, you move on to the next most likely issue.
Advantages of a hypothesis-driven approach to SEO
I’ve always been an advocate of thinking strategically and using data – and this approach supports both.
Here’s why starting with a hypothesis and testing it is, in my view, the best way to approach SEO:
- It keeps the scope manageable: You start with limited knowledge, so you’re not diving into every possible SEO issue. The search for answers becomes more focused and efficient.
- It avoids analysis paralysis: While you’re still using data (e.g., checking for broken links), you’re narrowing your attention to a few likely causes rather than trying to fix everything at once.
- It’s iterative: If the hypothesis doesn’t hold, you refine your thinking – drawing from case studies or others’ experiences to shape the next test – rather than blindly running through a checklist.
- It leads to real learning: This approach helps uncover meaningful insights, like: “If your Core Web Vitals scores dropped significantly during a migration, then fixing them can help recover traffic.”
Dig deeper: 5 practical SEO experiments with AI as a co-pilot
Progress requires a testing mindset
Adopting a testing mindset doesn’t just improve existing strategies – it opens the door to discovering new ones.
Take the example of a Bing employee who had an idea to change how ad headlines were displayed.
The idea was buried in a backlog and marked low priority – until an engineer ran an A/B test and uncovered a 12% lift in revenue. Without that test, the idea would’ve been lost.
Despite how quickly the SEO landscape is evolving, there’s still surprising resistance to new ideas.
It often seems that there’s only one way to do SEO, and it starts and ends with checklists.
Whatever your stance, the real question is: Have you tested it?
You don’t need to adopt every new concept, but you do need to evaluate them with an open mind.
Let’s not default to skepticism just because something’s unfamiliar. Let’s test more.
Most SEO teams still rely on checklists and guesswork. Learn how a hypothesis-driven approach leads to faster and smarter results.
