Going guerrilla — is it worth it?
Most product designers have used guerrilla research at some point in their career, whether they realize it or not. In short, guerrilla research is cutting standard researching corners and getting the minimum amount of information you need as quickly and frugally as possible. It could be pulling a friend to the side and asking them to walk through your prototype or interviewing strangers at a coffee shop.
When should you use guerrilla research?
A fast and cheap way to get user insights seems like every stakeholder’s dream. But, unfortunately, it isn’t as good as it sounds. Would you use duct tape to fix a car if you had access to a good mechanic? Probably not. It’s the same with guerrilla research. Duct tape is excellent for rough prototypes but not something we would want someone driving off the lot with. So before you hunt for users in the wilderness armed with your questions and prototype — consider the following.
What are you trying to find out?
If you’re looking for detailed insights, you may want to consider more structured user testing. If you’re going to get a broader pulse on your project to help move your work in the right direction, this is a good option.
When conducting guerrilla research, you shouldn’t come to the table with complex questions. This style of research lends itself to a more relaxed and conversational style of testing. In short, if you’re looking to:
- Explore the problem to better understand what you’re trying to solve
- Help guide your early design iterations
- Quickly debunk assumptions
Then congratulations! Guerrilla research might be right for you.
What phase of the project are you in?
There are different types of research and validation methods best for each phase of the design process. Some lend themselves across the process, and others are optimal for more specific use cases.
Guerrilla research works well in the discovery and exploration phase, especially in conjunction with other research methods.
This method is suitable for the discovery phase to better understand the problem(s) you’re trying to solve. However, this should be used in conjunction with other more in-depth research methods, particularly if there isn’t already data to go off of.
For the exploration phase, guerrilla research is acceptable to use as a stand-alone method to guide iterations and validate assumptions; this is the reason why guerrilla testing lends itself particularly well to agile environments. In addition, the fast nature of this method allows you to generally bypass the often lengthy setup and testing required by other more structured methodologies.
💡 Remember
Once you get to a more defined solution, you will want to move into more structured testing for detailed and specific insights.
How much time do you have?
Time is one of the most significant factors in deciding whether or not guerrilla research is for you. Do you have a reasonable amount of time on your hands? You may want to consider more structured testing to get more detailed insights. Were you supposed to solve a problem yesterday? Then, guerrilla research might be for you.
An important disclaimer here: If this is the only box you’re checking, I would highly recommend pumping the brakes and advocating for additional time for thorough research.
The only good bypassing best practices in favor of time will do for you is getting to a bad product faster.
How many resources are at your disposal?
If you scrape the bottom of the user testing barrel, then you’ll find good ‘ole guerrilla research. Having said that, if this is the only box you’re checking on this list, then you might want to push to have more resources dedicated to testing.
What happens if you cheap out on proper research and testing? Congratulations. You just paid for an expensive lousy product.
What are you trying to test?
It probably doesn’t have to be said, but what your testing is just as important as why you’re testing it when evaluating this method.
Does your testing need to be target-audience specific, aka, would an average Joe not be a viable test participant? Sorry, guerrilla testing isn’t for you.
Are you testing sensitive information, aka, will your client have problems if a random Joe who hasn’t signed an NDA sees it? Probably avoid this method.
Does your test revolve around a potentially uncomfortable subject? Will the questions you’re asking make an average Joe uncomfortable (think medical, political, financial apps)? Then you might want to reconsider.
🥳 Freebie alert!
Still don’t know if you should use guerrilla research? Take this quiz and find out.
So you checked all the boxes — what now?
Great. You’re ready to get started. Before you rush into interviewing strangers off the street, remember just because you’re roughing it doesn’t mean you can go in without a plan. The key to guerrilla testing is proper planning ahead of time. Before you start, you’ll need:
- Clear objectives
- A chosen method
- A discussion guide
- A plan of action — locations, target audience, approach
Setting your objectives
First and foremost, you need to understand what you want to achieve by running these tests. Your objectives will define the method you choose, how you shape your discussion guide, and what your plan of action eventually looks like. If you start without clear objectives, you’re going to be confused, and your test participants will be confused.
Start with the outcome
The best place to start defining objectives is by understanding what the desired outcome is.
Are you testing a paper prototype for a to-do list app? Maybe your desired outcome is to gain usability insights on how intuitive your “Create Task” flow is.
Are you exploring the problem space for a financial wellness app? Maybe your desired outcome is to understand how people think about student loan management.
Get specific and measurable
Once you understand the desired outcome, you can start to define your objectives. Remember: the more specific they are, the more successful you will be.
🚫 Example: A loosely defined objective
Understand how people think about student loans.
✅ Example: A clearly defined objective
To gain an understand of how people view their student loans specific to:
– Debt paydown
– Loan financing
– Financial wellness
The difference between these two statements isn’t just word count. It’s the difference between success and a wasted test.
Methods
Choosing the correct method will make a world of difference in the outcome of your test. There isn’t a one-size-fits-all, and what works in one scenario may not work well in another. The key to choosing the proper method is understanding your objectives and working within your constraints.
On-the-street interviews
On-the-street interviews are precisely what they sound like. It’s the process of flagging down random people in hopes that they will spare 5 minutes to answer a few questions. Depending on your approach and your location, these can be pretty successful. Although it sounds straightforward, knowing the best practices will set you up for success.
Pick the right location
Location is everything when it comes to performing impromptu interviews with strangers. Consider the following before deciding where to set up shop:
- Where would you find your target audience? If your target audience is people in their mid to late 40s, then going to your local college campus might not be a good option.
- Where and when would people be most receptive? A coffee shop at 1 pm probably isn’t going to be your best bet. Busy professionals stopping for a fuel stop on their lunch break probably won’t want to spare their time.
- Do you need space to show a prototype? Make sure you consider your environment before deciding where to run your interviews.
Give them context and make the ask
When you walk up to a potential test participant, make sure you approach with a smile and provide context. For example, something like the following should work well.
“Hey there! My name is ______, and I was wondering if you could spare a few minutes. I’m looking for feedback on an app I’m building.”
💡 Remember
If you can offer them a gift card or something similar as a value exchange. Do it.
Keep it brief
Nobody likes to have a “few minutes” turn into 15… then 20… and so on. Keep it brief. Try and limit it to no more than 3–5 questions per person.
Record or take notes for posterity
Ideally, if you can have a partner take notes, that would be best. Otherwise, be prepared to take notes on the fly or quickly jot everything down after your interview. The worst feeling is to feel like you wasted an interview because you can’t recall everything.
Remote testing
Remote testing is a bit trickier — it can easily cross the line into structured testing and lose the speed and cost-effectiveness of guerrilla testing. However, you can sign up for a free trial on someplace like usertesting.com or drop some money on running a short survey. If you need it to be free, then there is one option to consider.
Ask the internet
My favorite way to perform remote testing on a budget is to utilize the internet as a free resource — fun fact: people love to share their opinion. So if you want to better understand a problem space, ask open-ended questions about a topic, or get feedback on a prototype, then this is a great tool to consider.
My recommendation is to post on a forum such as Reddit.
💡 Remember
This is a double-edged sword. You have no way of validating they are who they say they are, so take their feedback with a grain of salt, especially if your test is audience-specific.
Building your discussion guide
Understanding what you want to ask before you ask is critical to success. Traditionally, a discussion guide is a set of tasks, questions, and topics that you’d like to cover over the course of your test. Since you’ll be keeping your interviews short and sweet, something simple and to the point will suffice. When building out your discussion guide, keep these things in mind:
- Set up your discussion guide with your objectives top of mind. You should set up the tasks and questions you ask to help you reach your objective.
- Ask open-ended questions. Asking leading questions can skew your test results with answers that may not be entirely truthful or accurate. Additionally, asking open-ended questions can also lead to more robust answers that provide more insight into why users feel the way they feel.
- Be willing to adapt. Your guide should be just that — a guide. This process is not prescriptive and if you happen to think of follow-up questions on the fly, go where the conversation leads you.
Creating your plan of action
You have your objectives, you’ve chosen your method, and you’ve nailed your discussion guide. You’re almost ready to start chasing down test participants. All you need now is a plan of action. Having a strategy is particularly important for in-person interviews; however, you should still have a rough plan of action for your digital guerrilla research as well.
Location, location, location
The first question you need to ask is, where will you be performing these tests? Consider your objectives, time available, and method to narrow down your location. I would recommend selecting a few different locations that fit within your parameters — this will allow you to cast a wider net and increase your chances of success. Remember to think outside of the box!
Are you not going to be boots-on-the-ground? You still need to consider where you will be conducting your tests. Are you going through a forum? Which one? Reddit? Which channel? Make sure the digital platforms you select cater to your target audience and support the test you are trying to conduct. Remember to abide by the rules of all the groups and channels that you decide to engage with!
Selecting your participants
Beggars can’t be choosers, but we can be smart about it. Unless you’re conducting a test that must target a specific audience (maybe you shouldn’t be using guerrilla research in that case), then try and cast a wide net. If you have the time, then it’s better to have more answers than not enough.
Remember: unless you’re a rockstar salesman, you will hear no a lot. So keep your spirits high and your charisma higher, and keep initiating conversations with people.
Selecting your approach — avoid being a car salesman
Choosing your approach is arguably the most challenging part of guerrilla testing. No one likes the mall kiosk people — so don’t be like that. Be friendly and approachable. Select participants that don’t look like they’re in a rush. In short, use common sense. Bonus: if you can, offer coffee or a voucher to increase your chances of success.
You got your data. What do you do with it?
You went out, asked the questions, and got the research. So before you run off to the stakeholders and let them know you should “implement solution A because Joe on Washington St said so,” — it’s time to compile the data.
I’ve found that the best way to do this is in a very structured and shareable format. Some best practices before you jump in:
- Keep it value-centric. How does your research tie back to business needs? How does it tie back to your objectives?
- Tie your test participants back to already established and familiar user personas.
- Have a mix of quantitative and qualitative data that support your assertion.
- Understand what you’re trying to achieve in sharing this information with stakeholders. Do you want buy-in for continued exploration in a specific feature? Do you want to educate on why you built something a certain way? Keep these in mind and tie the data back to these objectives.
These objectives may not directly relate to your test objectives. Often designer and stakeholder needs can differ, so make sure to tie these results back to your stakeholder’s needs for a better chance at getting buy-in.
In conclusion
Guerrilla research is like a spoon. It’s fantastic when you have a bowl of soup, but if you’re trying to eat a steak, then it’s going to cause more frustration than it solves. If you take anything away from this article, remember:
- Any research is better than no research. That, however, isn’t an excuse for not advocating for the proper research.
- Fast and frugal doesn’t mean easy or hasty. Successful guerrilla research requires prior planning.
- Document your findings in a way that’s honest and meaningful to your audience. Presenting your results to your fellow designers will often look drastically different than presenting your findings to stakeholders.
Ready to run your first guerrilla test? Good luck and happy researching!