#snapcast

Unless you've been living under a cave for the last week, you would've seen some form of Rebecca Black's music video, Friday. The video has garnered over 45 million views and has been a top trending topic on Twitter for the last 2 weeks.

Like all well adjusted internet users, I usually read comments when I am moved by a video, especially one with over 90% of dislikes. A lot of the commenters were incredibly passionate about their disdain for the song's inane lyrics, formulaic pop rhythm and the singer's auto-tuned voice. I didn't think anyone could have more haters than Justin Bieber.

As part of SnapCast's lean startup process, we've been doing problem/solution validation interviews for the last 3 weeks. Whenever we find someone who responses passionately about your problem or solution (good or bad), our eyes light up. We have spoken to many people who were luke warm - they were neither extremely excited nor incredibly negative. They say things like, "Yeah, I can see how this could work" or "If I tried it and I liked it, I'll use it". To me, reactions that are in the middle are the least useful when analyzing our data.

What you want to look for is a passionate "I love it" or "I hate it" response. Then you know you've hit a nerve. You want to avoid the dreaded "zone or mediocrity".

Obviously if you find users who are passionate about your solution, you are on the right track to turn them into your early adopters. But how do you deal with the haters? What if people hate your idea or your solution? How do you pivot? 

There was one segment (hobbyist photographers) we interviewed who were really critical of our idea and solution. (SnapCast creates visually engaging photo "stories" from your photo sets, sort of like a web version of Instagram, but for photo albums).  Another segment, mobile phone users, were really luke warm. We dug deep into our data and came out with very interesting insights, pointing us to a different segment of users. 

What’s the difference between and an “ideas” person and an entrepreneur? In the eyes of the lean startup movement, the “ideas” person is a dreamer with lots of unvalidated assumptions of how customers will use and buy their product. An entrepreneur knows that the initial idea is based on one a many hypotheses and that this idea is only the start of a journey of discovery. She must then test and validate these assumptions as fast as possible without wasting too much resources. Most likely, what product/service results at the end of this journey will be very different from the initial spark.   The prerequisites The latest idea we have is called SnapCast. We’ve gone through at least two different problem/solution iterations for SnapCast and several drafts of the lean canvas

We worked with Max Cameron (@maxcameron) with his Dooo.sh/it lean startup workshop to refine our canvas. From there, we laid out 3 main hypotheses that we wanted to validate. We decided that we would use interviews and ask very open ended questions to test our hypotheses. But our first tasks was really to recruit potential prospects. From our lean canvas we had identified 3 target segments.   1. Active users of photo sharing networks such as Instagram 2. Flickr groups related to photo collages such as diptic 3. Designers and creative types who are part of photo sharing networks The start - getting out of the building As soon as we had drawn up some questions, we immediately got out of the office and walked to the nearest Starbucks to recruit someone to interview. If you’re not a natural hustler or a salesperson, there is certainly a psychological barrier to “getting out of the building”. You procrastinate or make excuses not to talk to people. You need to break that barrier by just doing it and getting it over with immediately. We ended up meeting an acquittance of Max’s who is an avid Instagram user and interviewed him for about 40 mins. It was great, we had talked to someone and there was some sense of relief and excitement. We tried doing the same thing the next day, and failed miserably. Timing, we realized, was important. It was early afternoon, and the people there were either students, moms and their kids or the kinds of guys you see doing crossword puzzles at the subway. One way of building trust in a cold call recruitment at the local frozen yogurt place is to bring your well behaved kid along with you. (I try to expose my daughter to entrepreneurship at an early age) I haven’t tried it with a cute dog,  but I reckon it would work too.

The obvious While the coffee shop approach afforded us with immediate interviews and a chance to get over our fears and practice our questioning tactics, we had to come up with a systemic way to recruit prospects who were qualified. We started with the obvious - we knew people within our personal and professional networks who fit the target segments. Alan runs 2 popular photo sharing websites (myfourththirds.com and dpaddict.com) and has a list of thousands of serious photographers we could tap. We also asked our friends to help us recruit by sending them emails and asking them to forward to people they thought fit the bill. The interesting I personally wanted to recruit prospects in the following categories: extreme users, potential early adopters, subject matter experts and  what I termed as the “Gladwell social types”. I’m a big fan of Malcolm Gladwell’s Tipping Point and his observations on social epidemics. His “law of the few” states that “the success of any kind of social epidemic is heavily dependent on the involvement of people with a particular and rare set of social gifts." Gladwell called these people “connectors” (social butterflies), “mavens” (people you rely on for information) and “salesmen” (charismatic persuaders). My rationale was that if we could validate our hypotheses with people with these characteristics and later evolve them into early adopters, they could help us create our own tipping point for our product. Here are some examples of how I recruited for the following categories: Extreme user - I looked for a user with many posted photos and a very high rate of posting (>20/day). This person is literally posting a photo almost every hour. Potential early adopter - I looked for a popular user who had a twitter account and scanned his tweets to look for signs that would point to an early adoptoer. Connector - I looked for a user with the highest number of followers (keeping in mind the follower/following ratio should be > 1 as I discounted people who were following more people than they had followers) Mavens - I looked for users who were well respected in their community. In this case, I asked someone whom I personally admire and whose tweets/blogs I read. The process After we had identified who we wanted to ask, we simply contacted them and explained that we were doing research in the photo sharing space to validate a product idea and asked if they would like to help. I did a lot of “warm calls” - where even though I did not know a certain user, I would make sure I followed them, liked and commented on their photos and built some sort of rapport. I kept a spreadsheet of the users I wanted to contact and tracked them through the entire process. For example, on Instagram, I contacted 12 prospects, got about 7 replies and scheduled 3 interviews. A few who declined the interview were still ok to help, so I asked if they would prefer to do a survey instead. As part of this process, I had began to develop a network of prospective early adopters. Keeping it local We tried to get local prospects as far as possible. First, we preferred face to face interviews. Second, it was actually a lot easier to recruit people in the same city. Our acceptance rate was also way higher when we warm or cold call someone local. Next

We have a few interviews more to go through and we’ve also started to look at the data. Analyzing qualitative data is an interesting topic that I will deal with in detail in the next post.

Lessons learned

One of the most important principles of the lean startup methodology is to validate your assumptions during the problem/solution stage. This is where you have an idea and you need to figure out if someone will actually use/buy it. As part of our new project - SnapCast, we conducted over half a dozen interviews. What do I do with them now? I needed some framework to figure out all this raw data. So I asked a good friend of mine, Yileng Lee, who had studied at the Institute of Design and works as an senior innovation consultant for Doblin/Monitor.

My email to her was brief:  I need help from you about how to measure and evaluation qualitative data from the interviews. You are a pro at this. Do you have an article or a framework or something practical to follow... I want help to make sense of the interview data.

Her reply:

 OK, this is a bit of a longer question. Maybe should schedule some time to talk through it. In general it's messier to analyse qualitative data since you can't just tally it up and make a chart. 

High level:  Collect data  you can ask open-ended questions (e.g., "what do you do with your photos today?",  "how would you use this tool?" "for the things you said you would use this tool for, what if anything do you use to do those things today?" and/or probe around specific hypotheses "would you want to mash up photos? what things would you want to mash up? why?" )  Analyze  With qualitative,  you'll generally get a range of answers and diatribes that are a bit unwieldy to analyze. So you break them down into chunks. That's what post-it notes are for. Take each chunk of thing people said and put it on a post it note  -- e.g. "I like to create a collage from an event such as birthday party" That's analysis  Synthesize  Then look for patterns across all the post-it notes. That's starting to synthesize.  You can do a straight bubble up where you just look for patterns across all the post-it notes.  Or you can use a framework to initially sort the post-it notes into first. Generic framework might be AEIOU (activities, environments, interactions, objects, users)  In your case you can use your hypotheses as a framework. Sort the post it notes into each of the hypotheses. Which post-it notes relate to which hypothesis? (either supports or opposes it or influences it)  Interpret  Then interpret the patterns. What does it mean that there is this big cluster of comments about x or y.  What are the implications for Snapcast -- stay high level first instead of going directly to a solution. e.g., Snapcast should make it easier to identify which photos should go together