Persona Design Worksheet
About this resource
A persona is a reusable configuration that tells an assistant who it is acting as, who it is writing for, what it must always do, and what it must never do. Set one up properly and every person on a team gets consistent output without each of them having to prompt well, which is how organizations stop depending on their two strongest prompt writers.
This worksheet walks through the decisions: the role and its boundaries, the audience and their assumed knowledge, tone and register, non-negotiable constraints, what to do when a request falls outside scope, and how the persona should handle uncertainty. It ends with a testing section, because a persona that has not been tested against awkward inputs will behave unpredictably at exactly the wrong moment.
Configure it once and a whole team inherits it. This is how consistency stops depending on whether your best prompt writer happened to be the one who typed it.
How to use it
Complete it for one recurring task rather than for your organization in general. A persona built for a specific job with a specific audience works; a general-purpose persona is just a description of your company and changes nothing. Test it against the awkward inputs before rollout, because the moment a persona meets a request it was not designed for is the moment you find out whether it was designed at all.
Inside the download
What it is acting as, and just as importantly what it is not.
Who reads the output and what they already know, which shapes everything else.
Tone specified in terms of the audience rather than in adjectives.
What it must always include and never include, stated as rules.
Common questions
A reusable configuration specifying the role an assistant acts in, who the output is for, the tone required, the constraints it must respect, and what it should do with requests outside its scope. It is set once and applies to every conversation, rather than being restated each time.
Configure a shared persona rather than relying on each person to prompt well. Consistency problems are almost never a discipline issue; they come from every person supplying different context. A shared setup removes the variable, and it means a new team member gets good output on their first day.
Role and its boundaries, audience and their assumed knowledge, tone and register, non-negotiable constraints on what to always and never include, what to do with out-of-scope requests, and how to handle uncertainty. The last two are the most commonly omitted and the most consequential.
Free. No email list, no follow-up sequence.
More from the resource library, chosen because they pair well with this one.
How to use AI to write without losing the way you sound, plus the writing you should never delegate.
Eight reusable prompt structures that transfer to work you have not encountered yet.
How AI runs day to day once governance exists: routines, roles, decision rights, request and escalation paths, and the operating rhythm that keeps it working.
Tell us a little about yourself and this document is yours. We don't have an email list. We collect this so we know who's finding our resources useful, and that's it.
You're downloading
this document
Your information is kept confidential and never shared with anyone.
