Let’s Talk About Personas in the Context of SEO, a Hamsterdam Marketing Lesson

By Ethan Lazuk

Welcome to the first rendition of Hamsterdam Marketing!

So far in our Hamsterdam projects, we cover SEO history, current affairs, and AI research.

This new project will offer supporting context to all of these. We’ll be delving into marketing, branding, and related social sciences fundamentals.

In this first Hamsterdam Marketing lesson, we’ll discuss personas.

If you need a refresher, personas are fictional profiles that represent groups of similar people in your target audience.

I’ve noticed personas are generally referenced more for content marketing or paid media.

However, we can apply them to SEO, as well.

Let’s discuss how personas apply in the context of SEO, then we’ll review details about creating them.

How personas fit in the context of SEO

In SEO content research, having audience details from personas can inform the topics we cover as well as how we approach them.

For example, if we’re creating an ultimate guide post for a blog, we might include varying levels of detail or different subsections or perspectives based on the experience level of our audience.

Personas can influence other decisions around page experience or even content formats, too.

If an audience is older, for example, we might use larger fonts and shorter paragraphs to make a page easier on the eyes, or maybe we’d favor a desktop or tablet layout, if they’re more common with our audience.

On the other hand, if our audience is younger, we might consider embedding related social media content in our page, or even creating standalone social posts or videos that could surface in search results or AI overviews as “hidden gems.”

I will say, though, the traditional way I’ve seen personas created isn’t quite what we’re after for SEO. Even as I was doing research for this post, I saw a lot of good articles that nonetheless had these standard themes.

What I mean is, personas commonly have cute alliterative names, like “Couch Potato Carl,” and include bullet points with interests or demographic information that doesn’t always seem relevant for SEO strategies, or that can even feel stereotypical.

Take this example ChatGPT gave me (not sure why I chose slippers):

It’s interesting on the surface, but I would take it as a given that someone who wears slippers values “comfort,” and I’m not sure how applicable an interest in “mystery novels” would be.

Early in my SEO career, I thought personas weren’t very applicable because I largely saw them made this way.

I’d have rather just created content with focus keywords, where I could analyze SERPs and target search intents.

But I soon learned that while that style of content would be more likely to drive organic traffic, it still missed the mark.

Whether we’re targeting a keyword or making up a topic based on a list of hobbies, we’re still playing a guessing game to one degree or another.

Content that’s mainly keyword-focused has two disadvantages, in my opinion.

First, the way modern search engines evaluate content means they’re less likely to reward SEO tactics and are based more on finding patterns of user engagement and preferences with AI models, like neural networks. It’s unlikely a page that expressly targets a keyword without offering differentiated value (unique expertise or experience) can achieve sustained rankings.

Second, while some SEO clients might speak about rankings or traffic goals in onboarding meetings, what they ultimately judge success by is their business goals, like driving revenue from conversions.

I soon realized that in order to do well at SEO content, the audience needed to come first. That required a balance between targeting search intent and authentically solving their needs, like with insights from personas.

However, I still wasn’t getting the actionable insights I needed from the personas I had.

Then a watershed moment came when I first got to work with clients who were domain experts and also willingly participated in the SEO strategy and content creation processes.

Such domain experts could provide insights that not only differentiated the content but also provided relevant and actionable data about the target audience.

Take our slipper example. A domain expert might say something like, “Based on product reviews and customer service feedback, our target buyers care the most about price, sustainable materials, indoor/outdoor use, and a snug fit.”

Ok, now we have some context to work with.

Rather than plugging terms into content found through keyword research, we could address unique benefits on product pages or explore pain points in blog posts of known relevance to the audience.

But those insights weren’t always generalizable. Sometimes the clients had internal sales or customer service data, but other times not.

However, another lightbulb moment came when I got exposed to the data that email, paid search, and social media marketers gathered from their campaigns.

They knew, for example, which creative assets performed the best, or they could share audience demographics from large-scale datasets.

Simply following a persona spit out by ChatGPT might send you in the wrong direction or even offend your audience with inadvertent stereotypical assumptions.

On the other hand, grounding ChatGPT with expert insights or statistically significant email, ads, or social media campaign data could lead to more data-driven personas with interests and demographic information that could enrich SEO strategies with more holistic and nuanced context.

In fact, some of the most successful SEO content I’ve been a part of wasn’t based on keywords or competitive gap analyses but rather from incorporating insights from personas like these.

What makes a good persona?

Think with Google has a Future of Marketing section with an article from 2018 called Make it personal: Using marketing personas and empathy in your marketing.

It starts out by saying that trying to appeal to everyone puts you at risk of appealing to no one, and even well-defined audiences can still include many different types of people.

To avoid making “ineffective generalizations,” we can build audience personas that help us deliver “the right messages, offers, and products at the right time.”

One important point the article makes is that personas represent one similar group of people within your larger target audience, which means you’ll likely need multiple personas, and they could overlap.

Say we’re making decisions about the content or layout of a general service or product page. Those would likely apply to the entire target audience and require us to think about appealing to all of our personas.

This might involve us finding commonalities among the personas or incorporating different sections or page elements to address them individually.

On the other hand, content for more niche services or products might apply to only a few personas, in which case we’d want to cater to their needs.

The same is true for informational content. If we’re creating a guide for a medicine, for example, some patients might care more about prices or insurance eligibility, while others might be concerned with side effects or whether it’s appropriate for certain age groups.

Building personas

The most common mistake I’ve seen when creating personas is relying on assumptions from limited data or examples. This includes using default AI chatbot settings to create them.

The more proprietary and data-driven our personas are, the more accurate they’re apt to be, and the more insights we can find for our SEO strategies.

Collecting data

One place to start collecting data for personas is internally at a company.

Product reviews, email lists, CRM data, sales or customer service team data — as long as the appropriate permissions are granted, these sources can provide a wealth of insights.

This is where an LLM-based tool like ChatGPT (with its data analysis capabilities) can be most useful. We can upload spreadsheets or PDFs of audience information and build profiles from those foundations.

We can also collect data from customer interviews, surveys, or focus groups.

Email teams can be helpful in this way. They have reach into existing lists of customers and can also arrange campaigns to incentivize feedback.

The trick is getting a large enough sample size to make statistically accurate assumptions.

As a general rule of thumb (this is what Gemini told me, anyway), you’d need around 37 responses for every 1,000 people in your target audience to achieve a 95% confidence level.

You’d also want to beware of potential bias from selective response rates or the wrong sampling method.

Another source of large-scale datasets is your other marketing channels, like paid ads, email, or social media.

When I worked in marketing agencies, I saw a lot of siloing of different marketing channels. Each strategizing and reporting on their own efforts. However, cross-channel marketing efforts are generally more efficient and effective.

SEO teams can inform other channels of popular topics or audience pain points based on keyword insights, for example, while these teams can inform the SEO team about audience demographics or preferences for certain creative paid or organic media assets.

That data can help build statistically accurate data-driven audience personas, which can then be shared for a more holistic cross-channel marketing strategy.

Finding patterns

Once you’ve collected statistically significant amounts of data, the next step is to find patterns that are relevant to your specific audience.

You’re looking for dominant traits.

Your audience might be age agnostic, for example, in which case you wouldn’t care about building personas based on age groups so much as common interests or needs.

The Think with Google article suggests assigning a photo or fictional first name to your personas to make them more relatable, but I disagree.

In my opinion, by assigning human characteristics to our personas as individuals, we lose sight of the fact that we’re still targeting a collective and diverse group of people.

Even if we’re confident that a persona shares a dominant trait, like they all struggle with algebra homework, we don’t want to risk discussing our math tutoring service too specifically toward one person and risk alienating others.

I believe in following what I call an 11x content model. This is a collaborative approach that involves an SEO strategist, content specialist, and subject matter expert, where each of them imagines themselves in the position of the searcher and envisions the ideal solution. Then they work together on the final content.

When we create content for ourselves in this way, we tend to get more personalized, inspired, and higher-quality outputs that will resonate with our target audience as more naturally empathetic, at least in my experience.

Fictional people with made-up names and photos feel inauthentic to me, and I think that’s represented in the final results. That’s why I suggest keeping your personas data-focused while relying on your intuition for what feels most helpful and sincere.

Mapping the search journey

Once we have personas that are data-focused and centered around dominant traits, the next step is to envision their search journeys.

This is where we create a list of content topics and identify which personas they apply to, as well as what the best formats are.

For content to have SEO value, it should align with the target audience’s buyer’s journey. This means each piece of content solves a need they have while going through the funnel from awareness of their issue to considering solutions and then making a decision and eventually becoming a brand advocate.

The most common method for mapping this journey is to use keyword and competitor research. However, I think there are alternative sources of information that can lead to more original topics, like from reviews or social media content.

Especially in our current era of generative AI, where Google’s AI Overviews or AI summaries from Copilot or Perplexity can summarize basic information conveniently for users, we need more worthwhile SEO content that offers distinct expertise or unique experience that AI summaries can’t provide.

We should consider the best messenger for each topic. Maybe it’s helpful to get information in text or video form directly from a domain expert, or perhaps the audience would find it more relatable to hear a peer’s experience in a review, video testimonial, or social media post.

Secondary influences

Our target audience shouldn’t only include personas of people likely to purchase our product or service, but also people who can influence those decisions.

In B2B environments, for example, there may be higher-up decision makers who care less about a SaaS product’s features than their quarterly budgets or legal and privacy considerations.

It might make sense to speak to these secondary influences directly in some cases, or maybe address them through your primary persona, like by offering them tips on how to make the case for purchasing your product or service.

There might also need to be personas for detractors, or people who oppose your brand or product or service category.

Maybe you can address them and change hearts and minds, or maybe you want to avoid inflaming their passions by not drawing attention to any sore subjects, if you don’t have to.

On the flip side, keep your potential brand advocates in mind, as well.

Even if repeat purchases aren’t a common occurrence in your industry (say, for example, you’re a roofer or a surgeon), having personas for brand evangelists and incorporating them in your SEO and content strategy could lead to endorsements or referrals that bring new business.

Reevaluation

As you gather data from how your SEO strategy is performing based on your current audience personas, the last step is to reevaluate them.

I think it’s helpful to keep track of which content assets relate to which personas, so you can measure how that content is performing for different groups and spot any patterns.

You’ll want to go beyond keyword rankings and look at larger brand visibility trends (like impressions or mentions in AI summaries or chat experiences) and engagement metrics (like CTR, time on page, related page visits, or conversions involving organic search or other relevant channels).

See you next time!

I hope you’ve found this discussion around personas helpful.

This is only the first of our Hamsterdam Marketing lessons, with a new one to come next week.

I may also revisit this article to incorporate more information or visual examples, while this project itself will likely evolve as I learn what works best for you.

Until next time, enjoy the vibes:

Thanks for reading. Happy optimizing! 🙂