The job being hired
The progress each buyer is trying to make, written as a job they hire you for rather than a demographic bracket.
Free customer persona builder
Customer Personas
Paste your website and we'll read it, work out what you sell and who to, then build the personas most likely to buy - the job each one is hiring you for, what stops them, and the words that move them.
A persona is a job, not an age bracket.
Nobody wakes up and buys because they turned 34 and live in Brunswick. They buy because something happened. The kettle died. The old supplier missed a deadline. A doctor said a word that made them go home and search at 11pm.
That event, and the progress they are trying to make because of it, is the persona. Age, income and suburb are useful later for targeting. They are not the reason anyone buys, so they are a terrible place to start writing from.
Get the job right and the copy writes itself, because you already know the sentence they have been saying in their head.
The difference in practice
DEMOGRAPHIC PERSONA
"Sarah, 34, Melbourne, works in marketing, likes yoga and sustainable brands, income $95k, shops on her phone."
You cannot write a single ad from that.
JOB-BASED PERSONA
"Bought the cheap one, it failed inside a year, now refuses to be burnt twice and is looking for proof it lasts before she pays three times as much."
Now you know the headline, the objection, the proof to show, and the guarantee to put next to the button.
Why the old ones gather dust
Most persona decks are never opened twice.
Not because personas are useless, but because those ones cannot be acted on. If a document does not change a headline, an offer or a targeting decision, it was decoration.
- 01
The stock photo problem. A name, a face and a coffee order make it feel finished, so nobody notices there is no insight in it.
- 02
Averages describe nobody. Blend all your customers into one 34-year-old and you get a person who does not exist and would not buy.
- 03
It never names an objection. If the persona does not say what makes them hesitate, it cannot tell you what to fix on the page.
- 04
It ignores the trigger. Without the event that starts the search you have no idea when to spend, so you spend evenly and waste half of it.
- 05
It has no words in it. Real buyers use specific phrases. A persona with no quotes gives your copywriter nothing to work with.
- 06
Nobody owns it. Written once for a pitch, never tested, never updated, so the team quietly goes back to guessing.
What lands in the report
Every persona comes with the work attached.
Each persona is ranked by how much of your revenue it is likely to move, and says why it was ranked there. Then it gives you the eight things you need to actually run a campaign at them.
Trigger and timing
The event that starts the search, so you know when to be in front of them instead of shouting all year.
Push, pull, anxiety, habit
The four forces on the decision: what pushes them off the old way, what pulls them to you, what scares them, what keeps them stuck.
Objections and answers
The specific reasons they hesitate, each with the proof or copy that settles it.
Buying criteria
What they actually compare on, in their words, which is rarely the feature list you are proud of.
Where they look
The searches they type, the places they research, and the people they ask before they buy.
Messaging that lands
Lines that work on this persona, plus the lines that quietly repel them, with ad hooks and email subjects.
Targeting starters
A first pass at Meta and Google targeting for each persona, plus the proof your landing page needs to carry.
You also get an anti-persona, the buyer you should stop paying for, because cutting the wrong traffic usually beats adding more of it.
It reads first, then asks the right questions
A plumber and a SaaS do not have the same buyers, so they should not get the same template.
The tool reads your pages before it writes anything, works out what business model you are running, tells you what it based that on, and then changes the questions it asks. Same tool, different report.
D2C ecommerce
Personas split by the occasion and the price ceiling, and the report leans on repeat purchase and gifting.
B2B services
Personas gain a buying committee: the person with the pain, the person with the budget, and whoever can veto it.
SaaS
The job is framed around the workflow being replaced, and the report weighs trial friction and switching cost.
Local service
Urgency and trust dominate, so the report digs into what happens in the first phone call and who they ask first.
Marketplace
You get both sides: the buyer and the supplier, because a marketplace with one persona is half a strategy.
Nonprofit
Donors, volunteers and beneficiaries are separated out, since the money and the outcome sit with different people.
If your site is thin or the model is ambiguous, the report says so and lowers its own confidence rather than inventing a customer to fill the page.
Read this before you use it
These are hypotheses. That is the point.
This tool has never spoken to one of your customers. It reads your website, which is your version of the story, and reasons carefully from there. What comes out is a set of testable guesses, not findings.
Anyone who tells you a URL can produce validated personas is selling you something. What a good hypothesis does is save you weeks of staring at a blank page, and give you something specific enough to be proved wrong quickly. That is worth a lot more than a confident document that is quietly wrong for a year.
Every persona in the report ships with a "how to validate this" step, so you know exactly which question to ask a real customer to confirm or kill it.
Ring five customers
Ask what they were doing the day they decided to look. You are testing the trigger.
Read your own inbox
Support tickets and pre-sale questions are the objection list, already written by real people.
Run the hooks
Put two personas head to head in a small ad test. The market settles it faster than a meeting will.
Want these turned into campaigns?
Build the personas, then let our team test them properly: real customer interviews, the ads written to each hook, and landing pages that carry the proof each persona asks for.
Before you ask
Questions, answered.
They are built on the job someone is hiring you for, not on demographics. A persona that says "female, 35, urban, disposable income" tells you nothing you can write an ad from. A persona that says "her cheap one just broke for the third time and she has decided she is done" tells you the hook, the objection and the proof the page needs. Every persona here comes with the trigger, the forces pushing and pulling, the objections and the answer to each.
It cannot, not for certain, and the report says so on the first page. What a website does tell you is what is being sold, at what price, to whom it is written, which objections the FAQ answers, what the reviews emphasise and what the guarantees protect against. That is enough to build a strong hypothesis. These are proto-personas: a starting point to validate with real customers, not a substitute for talking to them.
It detects the business model first and writes to it. A B2B services buyer gets a decision process with other people in the room, a budget cycle and a business case. A direct-to-consumer shopper gets a moment, a price point and a returns worry. A local service buyer gets urgency and a suburb. Writing ecommerce personas for a consulting firm is the most common way these things become useless, so the tool checks before it writes.
A name and a one-line label, the job to be done, the trigger that starts the search, the four forces including what makes them hesitate, their goals and pains, three objections with the answer to each, their buying criteria and how the decision actually gets made, where they look for information, the exact phrases they type into Google, the messaging that lands and the messaging that repels, three ad hooks, two email subject lines, the proof your landing page needs, targeting ideas for Meta and Google, and the specific cheap thing you can do this month to confirm or kill the persona.
Three or four, ranked, with the reason each one sits where it does. You also get an anti-persona: who to avoid and how to spot them early. Most persona work skips the exclusion, which is why it never changes a budget decision.
Yes. The full set is on screen, with a print-ready version to save as a PDF and a Markdown file you can paste into a brief, a doc or your ad platform notes. No email required.
The report ends with the next steps, and they are deliberately small: read your last fifty support conversations and tag why each person bought, run five twenty-minute calls with recent buyers about what they did the day before they bought, then split your highest-spend ad set by persona angle and compare cost per purchase. That turns three hypotheses into one funded persona within a month.