A few days ago I hit the streets to interview random people for the dayjob. The company is big on design thinking and doing user interviews is one of it’s components. Ash Maurya (the author of Running Lean) refers this as problem interview whereas some other lean startup folks prefer phrase customer interview. Personally I feel that the phrase “problem interview” is better suited than “user interview” nor “customer interview” – since at this phase, those people aren’t your customers yet (they’re not paying you money) nor they’re your users (they’re not using your stuff).
We went around Singapore to interview the general public and met quite a diverse set of people who adequately represents Singapore’s demographic. Income levels ranges from housemaid to CFO with about half of them were Singaporean and the other half are on long-term residence visa. We deliberately try to avoid tourists since they’re not within our target demographic. After a day’s worth of running around we obtained 21 survey answers in total.
There were four of us and we split into two pairs. Each pair consists of an interviewer and a note taker. The guys acted as the note taker who just listened and took notes whereas the gals were the interviewers. We thought that people are more open to talk to ladies and thus how the team structure came to be (there is probably a reason why receptionists and stewardesses tend to be ladies and we didn’t have a strong enough case to go against this norm). Anyway here’s what you can learn from our experiences, without getting too much details.
Things to avoid
- Don’t chase people – if they walk fast, chances are the don’t want a conversation. Whenever we tried too hard and run after people, we always came out empty-handed.
- Don’t stand in lifts to catch people going out of of their residences. We did just this in the morning and hung around the void deck in a student residence apartment in a nearby university and everybody were in such a hurry.
- Office workers always feel they’re in a hurry during lunch hour, but it doesn’t always mean that they’re rushed – they just like to “think busy”. We saw a guy wearing office attire who looked leisurely shopping for mobile phone cases during lunch hour. When we asked for a short ten minutes he said he was busy. Probably busy worrying what kind of phone case that he’d like ^_^
- Don’t engage people who are on-the-job even though they look idle. This is in line with the “think busy” mindset above. We found a few staff of the commuter train company who looks like they’re not doing anything but they refused to be interviewed. We’re not sure why. But in any case there are plenty surveillance cameras around the stations and probably they don’t want to be caught on camera talking and instead prefer to be seen sitting around waiting for something.
- Don’t wear bank-like attire. The lady who played as interviewer in my team wore a blue blazer that could easily be mistaken as a credit card salesperson from Citibank or POSB. Whenever we were near banks or ATM areas we couldn’t approach anyone.
- Identify yourself clearly at the start and state that you’re not selling anything. People become defensive and raise their shields if they suspect you are trying to make a sale.
Things to do
- Try to find people who are alone because it’ll be a lot harder to single out an individual having a conversation with their group.
- People walking out of leisure places may be good interview targets. We got good results catching people walking out of a sports hall having done a few rounds of basketball.
- Smoking areas are a good interview spots – these are likely to be trash bins just outside office buildings. True that doing this is a workplace hazard since it makes you a passive smoker. But small stresses to your body should be helpful as what Nassim Nicholas Taleb said in AntiFragile.
- Long queues in taxi stands are also good interview spots. Sometimes it may take half an hour to get taxi at afternoon rush hours, as many people want to get home. Just make sure that there are at least about five other people (or distinct groups) waiting for taxis for a 10 minute questionnaire – typically it takes about two minutes per taxi when there is a queue in the taxi stand.
- Outdoor areas in coffee shops are another nice places to chat with random people. Primarily the ones that encourages you to linger. We got a good number of interviewees in the nearby Starbucks – just make sure that you avoid being noticed by waitresses or the establishment’s staff or risk them interrupting you.
- Keep the process short and within 10 minutes. Try to test out the interview guide first with your colleagues or friends and discard less important questions if it takes longer than ten minutes to go through them.
- Gain rapport first by asking general questions that doesn’t require much thinking and then gradually throw the more difficult questions. Ask demographic questions at the end of the interview and not at the beginning.
- It’s terribly hard to get office employee women ages 25—40 for an interview on the streets. They tend to act too busy. It was so bad and we got no answers from this demographic. However older female professionals tend to be more receptive and we interviewed a good number of them.
- People in the age of 25 to 50 tend to be hesitant expressing their age. However people outside that range always said their exact age – at least those we interviewed.
How we’d process the results
Processing the answers of open-ended survey questions are challenging. We settled by looking at each answer, deducing key concepts from them and counting those concepts. We didn’t want to lose out the unique and one off things and thus we list them down. Some of these one-off answers are gems and uncovered some key issues that we haven’t thought off before.
These survey answers helped confirm some initial of our assumptions of the core user problems. They also help in formulating personas needed for user story development down the line. These personas are reasonable amalgamations of the different clusters of common problems that a user may have. For example if person A have problems X and Y whereas person B are troubled by Y and Z and both A and B share significant demographic attributes, then we merge person A and B into one stereotypical persona.
Then the next step will be validation and prioritization – which user problems that are worth solving and in which order. This is the stage where business model canvas is in center stage. Each user problems are coupled with potential solutions and how a business model can back up those solutions. Naturally the user problems that can be backed by a strong business model makes good candidates to be solved commercially.
You might be wondering: why we interviewed just any people on the streets and haven’t done one more targeted to a specific market segment. Unfortunately I can’t disclose the content of the interview and likewise the reason for just grabbing anybody for an interview. When you work in a large corporation, you couldn’t say anything to the general public unless you’re an authorized spokeperson of the company or have gotten approval from the corporate communications department. That’s just part of the deal working with a megacorp.
I hope you found this useful. Until next time!