Persona Builder

The Persona Builder is a free, AI-powered buyer persona generator for B2B teams. Describe what you sell and who buys it, and it drafts one persona with goals, pains, likely objections, the channels they use, buying triggers and three message angles to test. The first time you use it, we ask for your email to show the results.

The output is AI-generated. It is a starting hypothesis, not research. The model has never spoken to your customers, so the draft must be checked against real conversations before you spend money on it.

It is for founders and small marketing teams who need a shared picture of the buyer before they write a landing page, plan ads or brief a sales hire.

Key takeaways

• The persona is an AI draft. Confirm or correct every line on real customer calls.
• Be specific in the brief. "Invoicing software for logistics firms, sold to finance managers" beats "B2B software".
• Use the pains as landing page headlines and the objections as FAQs and sales prep.
• Most B2B buyers aren't shopping right now, so the persona should shape brand messages as well as offers.
• One tested persona is worth more than five guessed ones.

How the persona builder works

You fill in two required fields and up to four optional ones:

Field	Required	Limit	Tip
What you sell	Yes	400 characters	Name the product and the problem it solves
Who buys it (role)	Yes	160 characters	The person who signs off, not “decision makers”
Their industry	No	120 characters	Narrower is better: “freight forwarding”, not “transport”
Company size	No	Pick a band	1 to 10, 11 to 50, 51 to 200, 201 to 1,000, or 1,000+
Market	No	120 characters	Country or region, such as US, Nigeria or UK
Anything else	No	400 characters	What they use today, who else is involved, deal size

Our server cleans the input and sends it to an AI model with fixed instructions. Those instructions ask for plain US English, no invented statistics, studies or real customer names, and a persona described “as a working hypothesis to validate with real customer calls”. The model must answer in a fixed structure, and our server checks it before you see it.

You get back:

• Persona name and role, such as “Finance Fiona, Finance Manager”.
• Summary: two or three sentences.
• Goals, Pains, Likely objections, Where they look for answers and Buying triggers: usually 3 to 5 short items each.
• Message angles to test: three one-line angles.

A note under the results repeats that the persona is AI-generated. Each visitor can run the AI tools 10 times a day.

How to read your persona

There is no score. The useful question for each line is: would a real buyer say this? We use a simple four-state check.

State	Meaning	What to do
Confirmed	Buyers said this, unprompted, on calls	Keep it and quote their words
Partly true	Buyers recognize it but describe it differently	Rewrite it in their language
Unknown	Nobody has mentioned it yet	Ask about it on the next call
Wrong	Buyers rejected it, or it never comes up	Delete it

Expect a mix. The draft saves you a blank page; the calls make it true.

A worked example

Brief: invoicing software for logistics companies, buyer finance manager, industry freight and logistics, size 51 to 200, market US.

Say the draft lists four pains:

1. Reconciling carrier invoices by hand.
2. Customers paying late.
3. Month-end close takes too long.
4. Hard to audit fuel surcharges.

You run five calls and count how many buyers mention each pain without being led:

Pain	Mentioned in	Share	Keep?
Manual reconciliation	4 of 5 calls	80%	Yes, lead with it
Late payments	3 of 5 calls	60%	Yes
Slow month-end close	1 of 5 calls	20%	No, for now
Fuel surcharge audits	0 of 5 calls	0%	Delete

Using a 60% bar (3 of 5 calls), two of the four AI-drafted pains survive. Say two buyers also raised one you didn’t have, such as chasing proof-of-delivery documents before invoicing. That is 2 of 5 calls, or 40%, so it goes on the list to ask about next time.

Two out of four is a normal hit rate for a first draft. The value is that you walked into five calls with sharp questions instead of a blank page.

Why message angles matter in B2B

• 95% of potential B2B buyers are out of market at any given time, according to the 95-5 rule LinkedIn B2B Institute, 95-5 Rule research with the Ehrenberg-Bass Institute (page checked 2026)
• 80% of companies change banking services once every 5 years, one of the purchase cycles behind the rule LinkedIn B2B Institute, 95-5 Rule research with the Ehrenberg-Bass Institute (page checked 2026)

The 95-5 rule is a rule of thumb, not a precise measurement, and the real share depends on how often your category is bought. But the arithmetic is useful. If you have 1,000 target accounts and roughly 5% are in market in a given quarter, about 50 are ready to talk. The other 950 still need to remember you when their trigger arrives.

That is why the persona’s buying triggers list matters. A new finance hire, a failed audit or a funding round are moments when an out-of-market buyer becomes an in-market one. Your messages for the 950 should build memory; your offers should be ready for the 50.

Turning a persona into action

From draft to working persona

• Run the tool with your most specific brief, then copy the persona into your team's notes.
• Mark every line Confirmed, Partly true, Unknown or Wrong as you hear from buyers.
• Rewrite confirmed pains in the buyer's exact words and test them as landing page headlines.
• Turn objections into FAQ entries and a one-page sales prep sheet.
• Use the channel list to decide where to test ads first, starting with one channel.
• Test the three message angles against each other in ads or email, one variable at a time.
• Review the persona every quarter, or whenever win rates change.

A persona sits inside a wider picture. Your target audience describes everyone you could serve; market segmentation splits it into groups worth different messages; the persona puts a face on the group you choose first. Our B2B lead generation guide shows how that choice flows into offers and channels.

Common mistakes

Myth vs fact

MythAn AI-generated persona is research about your market.

FactIt is a plausible draft built from your description and general patterns the model has seen. It can sound right and be wrong in the details that matter. Only real customer conversations, win-loss notes and data turn it into research.

Myth vs fact

MythThe more personas you have, the better your targeting.

FactEach persona needs its own messages, pages and tests. A small team with five untested personas usually ends up with generic copy for all of them. Start with the buyer who signs off.

Myth vs fact

MythDemographics such as age and hobbies make a B2B persona useful.

FactIn B2B, what matters is the job: the goals they're measured on, the problems that cost them time, who else signs off and what triggers a purchase. That is why the tool asks about role, industry and company size, not age.

Other mistakes we see often:

• Describing the product, not the problem. “AI-powered platform” gives the model nothing. “Cuts invoice reconciliation from days to hours” does.
• Naming a committee as the buyer. “Decision makers” isn’t a person. Pick the role whose budget pays.
• Never updating it. Markets shift. A persona from two years ago may describe someone who has changed jobs.

The persona I trust is the one sales has argued with. A draft from a tool, or from a workshop, is a set of guesses until someone checks it against real calls and lost deals. Start with the buyer who signs off, write down what you think they care about, then ask a handful of recent customers whether you got it right.

What the tool can’t do

It can’t interview your customers, read your CRM or know your win rates. It doesn’t check facts about your industry, and it can be confidently wrong. Use it to get from a blank page to good questions in a few minutes, then let your buyers correct it.