(Or anyone who wants to get involved)

Before I get started, my strongest advice is to hire or consult a UX professional at the earliest opportunity. They will help you plan, gather, analyse and apply user research, bringing your product to the next level much faster. Ideally, UX should be present from the very beginning at any startup that wants to succeed. 

Unfortunately, the reality is that a lot of Founders don’t have the resources to invest in UX, and they are often covering the gaps themselves. While I still think that doing some user research is better than doing none, there are a few hurdles to be aware of that might skew your results or give you a false sense of validation; This blog post provides you with a few basics to get you on the right track.

To provide a stronger UX foundation, I have invited former colleague Paulo Reis, Head of User Experience at Overleaf, to contribute his views and offer some practical tips (in italics throughout the post).

“User research is a deeply technical and scientific field of study. Survey design, conducting interviews, usability testing, qualitative and quantitative data analysis, synthesising and disseminating insights are all intricate enough that you can quite literally develop a career around them.

That said, user research is also fundamentally about empathising with your fellow humans. You shouldn’t need a PhD to get started. Talking to people (while being able to read between the lines), listening, observing, trying to elicit people’s thoughts without bias; those are all things that you have likely done in your personal and professional life. This post will help you employ these fundamentally human skills in a structured way to inform product development.”

Common scenarios

1. Your product is still in its infancy or doesn’t exist yet
2. You want to test a prototype to get feedback and iterate
3. You have product-market fit but want to discover a new problem space 

Research goals

As basic as it may sound, lots of people start researching without a clear goal; this makes the analysis of your data really difficult and can end up wasting your time. 

If you already have specific questions like “I wonder how much people would pay for this product” or “Is my product intuitive?”, go ahead and get started. Otherwise, block an hour in your calendar to make a list of the topics you need to learn more about – ideally work with someone else who knows the product. Your research topics could be anything from usability, to pricing, behaviour or context. Keep the focus on 1 topic, and formulate 3-5 very specific questions you want answers to. 

Tip: It could be tempting to generate a long list of questions while you have the opportunity, but that will only slow things down. Also, do not confuse your research questions (what you want to know) with what you are actually going to ask your users. These are very different! 

As important as your research questions, is the act of listing your own biases. Do you think you already know the answer to your research questions or have an approximate idea? Then record these as your biases and try to prove yourself wrong, you will often be surprised! This is a key step in any type of research, and it will give you much cleaner results.

Methodologies

Quantitative Research

Could you answer any of your research questions through data/analytics? For example, if your question is “How many people bought a subscription to a streaming service in the last year?” you may be able to find some data to answer that. This is called quantitative research.

If you are lacking data of your own, you can still look for relevant data elsewhere. There are plenty of industry reports or proxy sets of data that could still be helpful (hint: see what data is available about your competitors and you may have some answers!). 

Another way to gather quantitative data is to launch your own survey; even a simple one-question survey that requires a Y/N answer could really be helpful. Maybe you could have a pop-up on your website for a limited time until you get a statistically significant number of responses. 500+ responses are usually enough, but I’d aim for 1000+ if possible. Just remember to avoid leading questions and be aware of your biases. 

The great thing about surveys is that you can also ask respondents if they could be contacted for more in-depth interviews; You can ask them to leave their email address, or book themselves in via a calendar link. This is a very effective way to get your qualitative research started if you don’t know who to speak to.

Qualitative Research

If you are after a deeper understanding of user needs, desires or behaviours, opt for a qualitative research method. For example, if you want to know “What kind of value do people get from streaming services”, you could use one of the following methods:

• Ethnography
• User interviews 
• Usability tests
• Surveys

It won’t take you a lot of time or resources to get some meaningful data, you will be surprised how quickly you can start seeing patterns and get meaningful results. There are also some clever ways in which you can build these tools into your work routine. 

Ethnography

It may sound fancy, but all it means is that you observe users in their natural environment. For example, if your target audience are students you could spend a morning a month working from a library; don’t be afraid of approaching people, asking them questions or bringing your prototype with you. Think about places where your users hang out (this could also be online) and engage/observe those environments. Note how people behave, are they focused? Do they have a lot of interruptions? Do they have internet available? What devices are they using? Are there any barriers in their environment that could affect how they use your product? Keep notes of these insights and over time you will see some patterns forming. I was once surprised to learn that my users only had time to listen to podcasts, when I was trying to get them to read a paper journal!

Interviews

Interviews are a great way to discover unmet needs and get to the bottom of user pain points.

Whether in person or online, interviewing has three simple rules:

1. Use open questions, start them with “why” or “how”. For example, if you want to know about their experience with logging into an app, you could ask “tell me about the last app you logged into, how was your experience?”.
2. Summarise what the user has shared before moving to the next question – this builds trust and avoids misunderstandings.
3. Ask about behaviours rather than opinions. For example, “tell me about the last time you used AI” rather than “What do you think of AI?”. You will find your own answers as they describe their experiences, take some notes or record the session (with their permission). 

Paulo’s tip: If you have more specific questions to ask, leave them to the end of the session, after you’ve asked the open-ended ones — this will allow you to make sure you get the answers you need whilst still ensuring you get the benefits of discovery (i.e. understanding what’s relevant for the user, what they choose to emphasise). 

As for recruiting users, aim for 5-10 interviews to get a variety of insights. Keep your sessions short at about 20-30 min. If you are starting from scratch, don’t be afraid to talk to your immediate network and ask for introductions to people who may be close to your target audience. Speaking to someone is better than speaking to no-one! If your product already has a website or customer support, then these are great places to recruit. 

I am not a big fan of offering rewards for your research as they can really skew the results; users might feel pressured to tell you nice things about your product because they want the gift. If you have a budget to spend, then use products like usertesting.com where the infrastructure is already there and you can target different audiences. 

I also recommend setting up a calendar link where people can book themselves in, so you cut down on time spent setting up meetings. 

Usability Tests

The simplest way to run a usability test is to design a task for a user to complete and watch them as they do it – no interfering! There is software available to run these tests online in an unmoderated way (recommended), but nothing stops you from just putting your product/prototype in front of someone.

If you are present during the tests, be mindful that users may feel conscious of your presence and act differently because of that; also, if they know that you have built the product, they may be wary of giving you negative feedback. The analysis of these tests can be quite tricky, so I strongly recommend getting some help here if you can. 

Paulo’s advice: Sometimes doing some quick impromptu testing with whomever you can find nearby is better than spending hours speculating about a design on your own. Ask the user to think-aloud (i.e. explain their thought process as they go through the task); this gives you some extra insight and clarification at the expense of authenticity. A starting point for analysis (if you recorded the testing sessions) could be going through the recordings a first time to get a high-level understanding of the main usability issues (first-pass) and then doing a more thorough second-pass review, where you gather all issues, check which users are affected and tally them — this would give you a notion of the more prevalent issues, which typically should be addressed first.

Surveys

Surveys are relatively easy to set up and don’t cost anything to run; define a maximum of 5 questions, with multiple choice or Y/N answers. While having an open text box is nice if you want to gather extra insights, be aware that if you get a large number of responses you will need help/time to analyse the data, and often get poor quality data too. 

If you offer non-leading multiple choice questions instead, the analysis will be easy and the data will speak to you more directly. There are much more elaborate ways to run surveys but for the purposes of this blog post I recommend to stick to a simple approach.

Paulo’s tip: consider using standardised surveys (either as part of your survey or on their own). Standardised surveys have quite a few benefits: they have been crafted by experts, stood the test of time and often have benchmarking data available (so you can see how your product compares to others). Examples include the SUS (System Usability Scale), UMUX-Lite (Usability Metric for User Experience) or the CSAT (Customer Satisfaction Score).  

Analysis

In user research the outcomes are just as good (or bad!) as the research design. If you had a clear question, listed your biases, recruited the right users and kept the methodology simple, then it’s just a question of looking at the data against your research questions. 

You may want to visualise the results with a virtual board, or simple post-it notes. Look for themes, group things together and step away for a moment to think of the biggest opportunity. Try to avoid listening only to the loudest users and specific feature requests, instead think about the problems users are implicitly or explicitly telling you about (e.g. feature request: I want cake! vs. problem: I’m hungry!). 

Look again at your biases, did you prove yourself wrong (or right!) in any way? What was your biggest learning?

Wrapping up

I hope that this blog post was helpful in getting your head around some of the basics of user research that are most accessible to you. When you are finally ready to hire your first UX professional, look out for a well-rounded experience across qualitative and quantitative methodologies. If anything, you can use this blog post to prepare your interview questions!

Feel free to connect with me or Paulo Reis on LinkedIn to keep the conversation going!

If you are struggling for help, I offer a free 30 min call to discuss any product topic, including user research.

P.S. hit subscribe if you want to be alerted of new blog posts.