Meaningful Conversation 5: Building a Private Car Service Empire in 13 Minutes

I am running a social experiment to have more meaningful conversations. This is a recap of one of my attempts. If you want to see why I am doing this, please read this: My Experiment to Have More Meaningful Conversations.

This time I chose my cab driver.

I wondered what kind of ideas he would have for my iOS problem, but knew that first I needed to learn about his familiarity with his smartphone. We follow the same process when doing user testing for Clippo. Someone’s background shapes the questions you can ask. For a social media app, you’re going to ask a hardcore Instagram user different questions than someone who doesn’t use Facebook.

I started with a basic first check: what kind of smartphone did he have? Samsung. I knew he could not become an immediate user because our Android version has not launched yet.

Second check: what apps does he normally use? Text messages and maps. The fact that he told me texting was an app probably eliminated power user type questions.

Not expecting much, I told him how difficult it was for us to get users and asked if he had any ideas.

“I have not the first clue, but you seem like you know about the Internet – why don’t you help me.”

What can Walter White teach us about a private car business?

Now we were getting somewhere. He wanted to get more customers for his private car service in addition to those coming through using apps like Uber. It was his most profitable type of ride and he wanted to use the Internet to grow his customer base. I told him I would try to help, and started asking him questions to learn more about the numbers.

He told me how much he makes from each service, and I asked him what hours of the day and types of trips were the most profitable. The depth of his knowledge about the taxi business in SF was awesome – I learned so much.

After doing a quick analysis of his revenue streams, we decided the most profitable trip for him was to the airport using his private car service. What I told him was that his focus should be on booking as many of those kinds of trips as possible.

Then I told him that because he lives in the Bay Area, the Internet was probably the wrong way to get customers. For example, if he wanted to create an AdWords strategy for San Francisco, he would be competing against the savviest search marketing specialists from Uber, Lyft, etc. That’s tough for anyone, especially an AdWords beginner. What he could do, however, is try non-scalable tactics that would be hard for those companies to replicate.

Given his target customer was an airport traveler, we decided to focus on hotels. But, as opposed to going to large hotels, I told him he should try to form relationships with boutique hotels.

I think that boutique hotels gave him more opportunity for two main reasons: 1) Utilization – large hotels have lines of cabs waiting for passengers, so it was a fairly large opportunity cost to be sitting in the car hoping for an airport trip. 2) Competition – boutique hotels were less likely to be approached by rival cab drivers and would be more amenable to making him a preferred driver for a cut.

This gave us the strategy:

1. Create referral deals with the bellhops at boutique hotels for airport travelers. Why the bellhops? They are the most motivated to create a deal, because they can use the extra money, and are the ones that call the cabs. Also, skipping the hotel establishment allows the deal to close much more quickly.

2. In the down time between calls from bellhops, use Summon and Uber to fill the time and keep his utilization rate high. This would avoid long hotel lines and keep him earning while waiting for the higher payoff trips.

(Side Note: I had to leave Lyft off the list, because he refuses to drive for them. Apparently for a driver to sign up on Lyft, there is only a simple five-minute application. It upset him that inexperienced drivers could thus immediately be put on the road; he felt that he has paid his dues and needs to be treated differently. Though there is nothing he can do about it, I told him that I definitely agreed with him on principle. In fact, I think this lack of driver experience does make the Lyft product suffer a bit. When I took Lyft on Saturday, my driver did not know the city well (she was from San Jose) and we went the wrong way. I did not use Lyft on the way home. Step up your game Lyft!)

And that was that, we were able to create this plan in the 13 minute car ride. I gave him my contact info and I hope that he tries what we discussed.

Would you have added anything else?

Meaningful Conversation 4: Mongooses and User Experience

I am running a social experiment to have more meaningful conversations. This is a recap of one of my attempts. If you want to see why I am doing this, please read this: My Experiment to Have More Meaningful Conversations.

This is the first conversation in this series that is not centered on my work, but rather focuses on someone else’s work. I was in New York last week and one evening, I went to dinner with two friends and a girl I did not know.  At the table, she and I shared a side.

We started with small talk: the weather, what people did for work, etc. She mentioned that she was in the Columbia Anthroloplogy PhD program and I told her that I ran a startup. We moved on to cover food (we were at a pretty amazing Japanese restaurant), our trip to NYC, and other generalities. Since I am now constantly worried about falling into Type 2 Conversations, the basic trajectory of conversation worried me. So, I asked her more about her research. I told her that at a high level I knew what Anthropology was, but not really much more than that.

She started describing one of her former research projects, a study to preserve the seabird population in a specific habitat of Hawaii. Unfortunately to do that, their team had to control the mongoose population, which was eating many of the birds.

Animal rights issues aside, it fascinated me that someone was able to come to the decision that the seabird was more valuable than the mongoose. There are so many factors that determine an animal’s value in an ecosystem that I think it would be almost impossible to understand them all.

Rikki Tikki Tavi is the only mongoose I know

The thought process seemed unethical, until I found an analog in my life: controlling user behavior in a mobile app.

While not a predatory animal, any mobile app has predatory buttons, menus, and options that all compete for the user’s attention. When trying to determine how to have users take the actions we want, we also run experiments. Much like the researchers did with the mongoose, we have to predict the number one reason users are being led astray from what we feel is the optimal behavior.

For example, we just faced this with our new user tutorial. We created two versions, 1) have every new user make a Clippo and 2) have every new user just watch popular Clippos. We decided that having a user make a Clippo would make them understand the purpose of the app better. Then, we executed on that plan by actively restricting elements that competed for a user’s attention.

Though the teams had Anthropology and software development backgrounds, our experimental approaches were similar. I think the major difference is that unlike the mongoose, our users can choose to participate in the experiment. Then, to take that to the extreme, the mongoose lost its life in the process. This makes me a bit uncomfortable, but perhaps is the result of me not understanding the research industry.

What do you think?

Meaningful Conversation 3 – Seducing the Teller at Wells Fargo With Work

I am running a social experiment to have more meaningful conversations. This is a recap of one of my attempts. If you want to see why I am doing this, please read this: My Experiment to Have More Meaningful Conversations.

This was the first day that I approached someone that I did not know well to have a candid conversation. In fact, it was the first time that I actually interrupted someone’s life (and their job) to talk about mine. It was at the bank when I had to deposit some checks (I realize there are high-tech ways to do this, but I still take pleasure in depositing them myself).

I decided beforehand to talk to any random teller, whether I knew them or not. I did not expect to be nervous, but I was. I had the same feeling that I get in a bar before approaching a pretty girl. The name of this blog is ‘Shooters Shoot’ for a reason, so I just started with, “I’m excited to show you the new app we just released.” The teller was excited and so was the manager next to him. I showed them Clippo and they thought it was cool, but that was not enough alone to sustain a longer conversation.


The manager started to walk away, so I explained to the teller loudly enough so the manager could hear, how we were having trouble getting users and asked if he had any ideas. The teller said, “what about Facebook?” and then there was another 30-second silence. It would have been easy to give up there, but I kept pressing. I described how acquiring a mobile user on Facebook was quite expensive and it was only for larger companies like gaming companies with deep pockets who could buy users.

Thankfully, this brought the manager back (probably wondering what I was still doing there), with an idea. “You know,” he started, “one of my friends built this photo app that is one of the top 10 on the app store.”

Then the conversation started rolling. In fact, I ended up talking to them for almost fifteen minutes, completely monopolizing this teller’s time. What was even better was that he is an avid Vine user, so I was speaking to our target demographic.

Reflecting on this chat, I found two things interesting. First, I realized that there was a large barrier that I needed to surpass to even have this conversation; the teller needed tacit “approval” from his manager to talk to me. Second, once we started talking more personally about my problem, and not just the app itself, the teller began to think of more ideas, and became invested in the conversation. Maybe this is why as a startup we are supposed to “marry a problem” and not the solution?

Again like my first encounter, there was a fair amount of give and take. They offered ideas for my user acquisition problem and also asked some questions about startups, mobile apps, and other tech-related things they were curious about. Perhaps they didn’t have an outlet for this type of conversation or maybe they were just being polite and humoring me, but regardless I was happy to answer their questions. To be honest, I’m excited about taking my next trip to the bank- I am curious if our conversation has the potential to go deeper.

In addition to my commentary, I’m going to add a summary at the end of each post, so I can keep track of what the other person did, what they shared, and what I answered for them (if anything).

What they did:

  1. They both downloaded the app (2 new users!)
  2. The manager told me about his friend who made one of the 10 most popular photography apps in the app store – and one of his early user strategies (it’s a good one)
  3. One of them gave me feedback about the tutorial
  4. Both said they would try it and share with friends and will have feedback the next time I come in (I can’t wait)

What they asked me:

  1. What do you keep track of in a new app?
  2. How many users do you need to have?
  3. What was easier, building a product or finding investors? (I thought this was an interesting question – they didn’t ask about acquiring new users)

Meaningful Conversation 2 – Talking to Family About Your Startup

I am running a social experiment to have more meaningful conversations. This is a recap of one of my attempts. If you want to see why I am doing this, please read this: My Experiment to Have More Meaningful Conversations.

I am genuinely excited by my pseudo-experiment (I don’t think I can officially call it an experiment because I have no real hypothesis). After my first success, I was ready for attempt number two. As opposed to a semi-stranger, I decided to play it safe and try my sister.

This is such an effective chart.

I want to include family because, for me, it’s always been tough to talk about Clippo with them. It’s their job to be worried about my well-being and since startups are a never-ending series of volatile emotional swings, it’s easier to just talk about something else.

I feel like I can already hear my psychologist roommate saying this is the exactly WRONG way to handle your emotions. My naïve method was to just email positive press and give high level updates. I am not really sure how I ever thought this was a good idea. I started doing it a few years ago when we were bootstrapping the company and I was working out of my apartment and living with my co-founder. I was having a real issue separating work from the rest of my life.

I wouldn’t be surprised if this is common for those that work from home, since there is no physical separation between work and home life. I created a plan to have designated “work” times and designated “no work” times. Admittedly, this plan does not work very well–I am always thinking about what products we are building and how to make them better. However, in my mind, having some separation, even an imaginary one, has become very important to me.

My conversations with family members easily fall into the “no work” category because there are so many non-work things to talk about. But to be honest, I really just don’t want to worry about them worrying about me.

The goal is to remain one big happy family.

This time, when, my sister asked, “How is Clippo going?” I came clean. I told her we had just launched a major bug-squashing update to the app and that our largest problem was figuring out how to acquire new users and bring back our old users.

I think my sister skews my experiment results for two reasons: 1) she is family, so is going to go out of her way to help me, 2) she knows tech and marketing – she worked in the industry for about 8 years and is currently a rising star at a very successful start-up. I was not exactly asking a random stranger on the street for user acquisition advice.

She had some really great ideas. My favorite was the #getweird campaign. We were trying to brainstorm a hashtag that we could “own” and would spark people’s creativity to make funny and shocking memes. The idea is to have people make Clippos of the weirdest things they see on the street, at work, late at night, or at home. (In fact, I’m going to be running around San Francisco next week making funny memes in offices, on the streets, everywhere. If you are interested in helping out, please email me.) I would definitely count this as a positive outcome for experiment trial #2.

Going 2 for 2 so far has my confidence about as high as it can get. After sharing drafts of this post with a few people, they asked some interesting questions that I do not have the answer to yet. I thought the best way would be to share them below. Hopefully, I’ll be able to better answer these in future posts.

  • Are you worried how conversations like this impact your personal work/life balance?
  • Are you going to reach out to her more regularly?
  • Do you think these conversations will actually improve your relationship with her?

I think the next iteration for family is to talk to a non-tech knowledgeable family member and see if the result is as positive. I’m curious for anyone reading this – do you have problems talking to your family about your work?

Thanks to Vineeta Agarwala, Nick Meyer, Daniel Post, and Ritika Goyal for reading.

My Experiment to Have More Meaningful Conversations


My typical day is filled with 3 kinds of conversations.

  1. 2-10 seconds. I like to call this a “freshmen dorm conversation.” I’m now five years removed from college so I’m not sure if I’m still allowed to make college analogies but here it goes: you see someone who you lived with in a freshman dorm and haven’t seen since. You say hello, and the conversation follows a set pattern: “Hey, how’s it going?” They reply “Good, how are you?” and you move on.
  2. 30 seconds – 3 minutes. This is someone who you see on a semi-regular basis and you start the conversation with general questions and maybe have time for a weather-related or current events topic. Then, you close up shop and move on.
  3. 3 minutes+. This does not seem like a lot of time, but how many of these types of conversations do you really have? To keep a conversation going longer than three minutes there has to be mutual interest and at least one meaningful topic of discussion.

I generally tend to have Type 2 conversations. This could be because I help run a startup and meet a lot of startup people; everyone I meet has his or her elevator pitch or talking points well rehearsed. In fact, I have a lot of conversations that are abridged versions of “the show,” which to me means anyone describing his or her startup with a positive spin.

I don’t know about you, but to me these Type 2 conversations are frustrating – especially when talking to interesting, intelligent people working on startups. It’s not really a secret that every startup is horribly broken anyways.

Entrepreneurs are eternal optimists, so I understand the impulse to use a positive word shield to hide what is really going on, but I wonder how helpful people would actually be if you let them try. For example, whenever we had office hours at Y Combinator, the partners would always ask what our most pressing problem was, and those sessions were incredibly productive.  This made me wonder, what if I did this with people who I see every day?

After having any idea, I run it by my “Am I crazy?” filter: my mom. She lives in a no B.S. world and is the perfect first person to run any idea by. Her response, “you are probably crazy. People don’t have time for your problems.” Since both of these are probably true, we created an important rule for a potential experiment. Don’t tell people my most personal problems, yet. Until I have a better idea whether this works or not, my shared problems will strictly be work related.

Like any supporter of the lean startup philosophy, before talking to anyone else or even writing this post, I decided to test this idea once. I’ve been going to the gym pretty regularly for the past two years and there is a guy who I see there a few times a week. We’ve probably had hundreds of Type 1 conversations and at least twenty Type 2 conversations. I still didn’t know his name or where he worked. He was a nice, anonymous gym guy.

This time, however, when he asked “How’s work?” instead of responding “Going great,” I replied that we had just launched our new app, Clippo, in the app store and were having a real problem with our user acquisition strategy. Namely, that paid apps and gaming companies had driven user acquisition costs so high that it made it very difficult for a consumer product without an extremely high budget to be competitive.

Gym guy didn’t respond with “that’s tough, good luck,” which is what I half-expected. In fact, he explained how the company that he works for created an optimized system for AdWords that would help us reduce costs and create a framework for user acquisition.

Some day I’ll have the lyrical flow of Eminem.

To quote Eminem, “Ah wait, no way, you’re kidding. He didn’t just say what I think he did, did he?” An insight and a strategy in my first Type 3 conversion experiment!

I made sure that I really understood what he had just explained to me. I asked at least five to ten follow-up questions, and gym guy answered each one patiently. I could tell he really enjoyed talking about his strategy and was excited to share it with someone who he knew was obviously grateful. How did I know that? I asked him.

I actually will be implementing his Adwords strategy over the next few weeks and, depending how it works for us (his company is in ed tech), will potentially write a post about it.

But my favorite part came after that – he told me one of his problems. For privacy sake, I won’t share what it was but am happy to report that I had a couple ways that I thought might help him. By approaching him with my problem, I think it made him comfortable enough to share something with me. Again, it is important to remember that before this we had never had a conversation for longer than 2 minutes. I wouldn’t say we are friends now, but I do know his name.

That exchange is what gives me real confidence that this can work. Another question my mom brought up was, “Is the average person going to know or care about your startup problems?” At first I had no answer, then I came back thinking “probably not and that is a good thing”. One thing we struggle with from time to time is uniformity of ideas. Christian and I work so well together and think so similarly that sometimes we are too close and too involved.

I’d love for some of my friends or anyone, maybe someone who works in finance or materials research, to tell me how they think I should acquire users. Maybe it will be awkward and they won’t care or have ideas, but maybe they will come up with something or some accessible group that I would have never considered. That to me is an exciting outcome.

This brings me to my plan: try and take one Type 2 conversation and move it to a Type 3 conversation every day. (My philosophy to stick to a plan is to take them one small step at a time.) Then, in every one of these conversations, I am going to lead with my most pressing work problem and write a post about how it went.

If the first couple weeks go well, I’m going to add personal problems to the mix. If it’s a disaster, I can always record them, make a reality show, and drive viewer growth using strategic Adwords placement. Someone asked me what my goal for this experiment is and honestly I don’t have something I’m trying to learn. I just want to see if it works for me. Am I missing something, how do you drive meaningful conversation in your life?

Big thanks to Nick Meyer, Melanie Major, and Sheila Vashee for editing.