I'm in the position where I can not do proper user research for an upcoming project. I do have some information from the customer (their RFI and RFQ) but I don't feel that is really enough.

How would you go about creating personas and user profiles? Is there some way or process to make more educated guesses?

8 Answers 8


What might be constructive is to try to create personas from the information you do have.

This will highlight the assumptions you are making and the information you don't have.

You can then use this to go back to the customer with specific questions rather than having general "tell me what you want" conversations.

  • 2
    For future users I would like to highlight that doing this simply isn't possible in terms of the process that creating personas requires which is real primary data. Without primary data you can not create personas, then they are called "ad-hoc personas" or "user profiles". They are often created as a temporary step to understnad which questions to ask during the research for creating real personas. Apr 18, 2013 at 20:52

If you can't find real users to talk to, it can help to talk to some of the people who do talk to real users - or have a job that means that they have to empathise with them. People like:

  • product managers
  • technical authors
  • customer support
  • sales

can often give useful insight.

  • Particularly customer support !
    – PhillipW
    Nov 21, 2013 at 10:37

you can use Tamara Adlin's technique which called Ad-Hoc Personas. "Ad-hoc personas = testable hypotheses. It will give you a place to start gathering more data. This process will give you roles and goals which is the first step to developing personas."

Read this great article for more information: http://www.thingsontop.com/remixing-power-adhoc-personas-1020.html

Good luck,

Tali Hirsh


There is nothing wrong with going back to the customer for more information. They have a vested interest in seeing you succeed. If you're having difficulty, put your request to them in business terms. Something like:

Dear ____,

I'm very excited to be working on X and I appreciate the opportunity to do so. Because I value the investment you've made to see this project succeed, I want to put forth my best possible effort.

While the information you've given me so far is helpful, the success of the project and its return on investment would greatly be enhanced with more information about who will be using X, why they will be using it, and what benefits they expect to receive from X.

Specifically, I need to know A, B, and C. The reasons I need A, B, and C are D, E, and F.

It also helps me to understand more about your business goals for project X. What specific actions are important for users of X to perform in terms of your business?

The most successful designs are when business goals and user goals overlap. It's my job to design X to do that, so every detail is vital to make X as successful as possible. I find that it's best to do research in the beginning as it becomes very time-consuming and expensive to make changes later on.

Thank you for your time.




  • If you're at the stage of dealing with an RFI and an RFQ then this kind of dialog isn't going to be possible. Good advice for later in the process, but not possible up front.
    – Bevan
    Oct 10, 2010 at 23:06

I’m assuming that what you mean is that you have time and other resources to do research, but don’t have direct access to the users in order to conduct observations, interviews, or surveys. There are a number of methods of user data collection that depend on archival sources rather than direct access:

For users of a corporate or business product, you can ask your client if you could see:

  • Job descriptions and organization charts to understand the responsibilities (i.e., goals) of the users.

  • Personnel records or reports for extracting the experience and knowledge of the users.

  • Procedure manuals and policy statements to understand the task and information flows (at least how it’s “supposed to be done”).

  • Inventory of tools, software, and sites currently used (or at least given) to the users, along with access to the same (or at least to their user manuals) to see what users are used to (e.g., information and menu organizations; symbols, codes, units, formats, terms, abbreviations, and layouts).

For users of a consumer product, you can ask your client for:

  • Business plans and marketing strategies relevant to the product, to identify the ultimate goal or position of the product.

  • Market research results, to identify characteristics about the user and the specific needs the product is aiming to fulfill.

  • List of existing competitor products for study, to identify what users are used to (or at least inclined towards), and also to determine what can be improved in the UX (by cognitive walkthrough or heuristic analysis).

For either, ask for:

  • Hit logs and database entries and reports for the current system to see what users are currently doing (e.g., how often they engage in each activity, what sort of entry errors they make).

  • Records or reports from tech support, complaint departments, customer service, and critical incidents for identifying problems with the current system.

Of course, nothing beats being face-to-face with users to see what really matters to them. But the above methods are good as a substitute or at least a supplement to limited direct user access.

I’d be careful about making a persona too detailed or refined if you have limited or unreliable user data. Rich personas can be very compelling, which would give your product team members an illusion of greater accuracy than there really is. In this case, you may want to choose a different way to summarize your research findings that highlights the uncertainties to your team.


I recently came across a technique called Proto-Personas in Leah Buley's The User Experience Team of One. This is a formalized technique that mirrors others' advice above (provisional personas, ad-hoc personas). Different names. Same general idea, I think.

Leah Buley:

They [proto-personas] can be created with the help of the cross-functional team... In essence, proto-personas are a persona hack that you create using whatever data you have available and with the help of the team.... [A] proto-persona is based on whatever insights you have, which can include secondhand research, or even the well-informed hunches of a team of people.... Less scientific and rigorous than traditional personas, but they can be equally effective for helping a team shift into a more empathetic mindset.

Disclosure: I'm not at all affiliated with the author or publisher. I just found the exercise she describes useful.

You can find details here: https://rosenfeldmedia.com/books/the-user-experience-team-of-one/; or the Amazon link here: https://www.amazon.com/User-Experience-Team-One-Research/dp/1933820187


You will need to develop provisional personas based on the knowledge of stakeholders within your company, and any secondary research you can do. I developed a presentation to aid in the creation of personas without research that I think might help: http://slowmtn.tumblr.com/post/76450958623/provisional-persona-workshop


Here's a suggestion:

You could write up your best description of personas based on people you know (in your company, friends, relatives, etc.) and then iterate with the client to fine tune it. It also helps to write out user scenarios to test out the behavior of the personas. Scenarios help you determine whether your assumption would work. You don't have to get it perfect the first time. Iteration is part of design. It will help make it better.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.