In kickoff meetings with new clients I always ask to see their buyer personas, if they exist at all. I have been handed true works of art: profiles with a stock photo, an alliterative name like "Marketing Mary", age, neighborhood, pet and favorite TV show. And when I ask what decision they made differently because of that profile, the most common answer is a long silence. In many companies the buyer persona ended up as a decorative deliverable, a slide that gets presented once and never opened again. Which is a shame, because when it is built properly it is one of the most practical tools I know for deciding what content to create, what to say on each channel and which searches to go after.
In this guide I will explain what a buyer persona is, how it differs from a target audience, how to create one with real data (and not with a brainstorming session in a meeting room) and how to use it in practice, including the part closest to my own work: content and SEO.
What is a buyer persona?
A buyer persona is a semi-fictional portrait of your ideal customer, built from real data: interviews, sales history, analytics, reviews and searches. Semi-fictional because the character as such does not exist, you invent a name and a face so the team can talk about them. And data-based because everything that matters in the profile, the problem that brings them in, what holds them back, where they get informed, comes from patterns observed in real customers.
What makes the format work is that it condenses an entire segment into one person the team can reason with. Writing a sales page for "mid-sized logistics companies" is hard, because you cannot guess the doubts of an abstract group. Writing it for "Andrés, head of operations, who needs to justify every expense to a skeptical board" is a different story: you can ask yourself whether the headline would mean anything to him, whether the price will create an internal problem for him, whether the case study you picked resembles his reality.
How is it different from a target audience?
The target audience is the group: professionals between 30 and 50, in a given industry, with a given role, in a given city. The buyer persona takes one representative person from that group and looks at them up close: what problem brings them in, who they ask for recommendations, what makes them hesitate before signing, what they search on Google on a Tuesday at eleven at night.
The two concepts work together and in a specific order. First you run a market segmentation to divide your market into groups that buy differently. Then you choose which segments you are going to serve. And only then do you build one or two buyer personas per priority segment. A persona without prior segmentation is a character floating in the air, with no market size and no assigned priority. And a segmentation without personas usually stays in a spreadsheet that the creative team does not know how to turn into headlines.
What is a buyer persona for?
For making at least three concrete decisions. It is worth naming them, because this is exactly where the useful tool parts ways with the decoration.
The first is what content to create. Each persona asks different questions at each stage of the sales funnel, and the profile tells you which ones. If you know Andrés needs to convince his board, what helps him is less another generic benefits post and more a return calculator or a documented case with numbers he can forward internally.
The second is how to talk. Vocabulary, tone and objections change by persona. A finance manager wants to see the risk covered before hearing the promise, while a startup founder wants speed and tolerates mistakes. The same product gets presented differently to each one, and the profile works as a permanent reminder of those differences.
The third is where to invest. If your personas get their information on LinkedIn and in niche newsletters, the budget you keep in other channels needs a good reason to stay there.
And what is it not for?
For turning buyers away. The persona prioritizes your marketing efforts, and if someone who does not match the profile buys from you, they are welcome. It also should not be asked for census-level precision, because it is a tool for focusing decisions and it will always simplify reality.
How to create a buyer persona step by step
1. Squeeze the data you already have
Sales history, CRM, site analytics, WhatsApp conversations, reviews, support tickets and the sales team's memory, which tends to be the cheapest market research available. Guiding questions: who are the customers that bring in the most revenue, which ones reached the purchase fastest, what do the ones who left have in common. With that you can already sketch two or three tentative profiles.
2. Interview real customers
This is the step that separates a persona from an invented profile, and with five to seven interviews per profile patterns start to show up. The questions that have given me the most information: what was going on at your job when you started looking for a solution, which alternatives did you compare, what almost made you not buy, and how would you describe the problem to a colleague. Save verbatim quotes. The exact way a customer describes their problem is gold for headlines, for ads and, as we will see, for SEO.
3. Add search data
Here comes my favorite part, because searches complement interviews in a very convenient way: interviews give you depth with few cases, and searches give you volume with thousands. The keywords your market uses show how people describe their problems when nobody is watching, which questions they ask and in what quantity. If you have Google Search Console, the queries that bring people to your site are a free x-ray of which personas are finding you today. And the related questions Google shows around each topic hand you the map of doubts for each profile.
An example with invented numbers to illustrate the mechanism: if "software for clinics" gets 900 searches a month and "how to comply with health data regulations" gets 400, those are two different personas, one evaluating tools and one with a regulatory problem who will probably end up buying the same thing. Each one needs its own content.
4. Build a one-page profile
A name and one line of context, the goal they are chasing, what holds them back, where they get informed, their most frequent objections, one or two verbatim quotes from the interviews and, most importantly, a section called "what this means for our marketing". If a field does not change any decision, it does not belong. Mary's favorite TV show has never improved a campaign.
5. Use it and keep it updated
The profile has to show up in the brief of every campaign and at the base of your content plan. And it has an expiration date: markets move, roles change priorities, and a three-year-old persona describes a buyer who may no longer exist. Reviewing it once a year is a reasonable minimum.
How many buyer personas do you need?
Fewer than you think. For most businesses, one to three personas cover most of the revenue, and my favorite test is simple: if nobody on the team can describe them from memory, there are too many. Some companies also add a "negative persona", the profile that consumes support and discounts without leaving any margin, so they can recognize it early and stop chasing it with campaigns. It is useful, as long as it does not become an excuse to treat someone who already bought from you poorly.
Which mistakes should you avoid?
The first and most common: building the persona in an internal meeting, without talking to a single customer. The result is corporate fan fiction, a character that confirms everything the company already believed about itself. The second is filling the profile with decorative details, hobbies, favorite movie, car brand, that do not change any marketing decision. The third is portraying the customer you would like to have instead of the one who actually buys from you, which amounts to the aspirational version of the first mistake. And the fourth is leaving the profile frozen while the market moves.
What changes with AI assistants?
Quite a bit more than it seems. When someone writes to ChatGPT "I run a dental clinic with three locations, patients stop coming back after the first visit and I do not know whether I need software or an agency", they are handing over their complete buyer persona profile in one sentence: context, problem and the alternatives they are weighing. The model responds with recommendations tailored to that profile. On top of the classic results page, your content now competes inside conversations where each user describes themselves in full detail, and content that answers to specific personas, with their context and their objections, has better odds of being cited in those answers than the generic page that talks to everyone.
The other side is that AI also helps you build and use personas. It is good at synthesizing interview transcripts and finding patterns that slip by when you read them one by one. And one step further, you can turn the profile into a GPT you talk to before publishing: you show it a headline and ask whether it would care. I wrote a complete guide on how to create a persona GPT if you want to try it.
Wrapping up
This week's exercise needs no budget. Schedule three twenty-minute conversations with current customers and ask them the four questions from step 2, saving verbatim quotes. Then open Google Search Console, take the top 50 queries that bring people to your site and group them by the need behind them. With those two sources, write a single one-page profile, the one for the segment that brings in the most revenue. And apply the final test: if that profile does not make you change anything in the content you had planned for next month, it still lacks data. When it forces you to drop a weak idea and add two better ones, it is working.