The app that wouldn't go away

For an app that’s all about disappearing content, Snapchat just wouldn’t leave me alone. The routine went something like this:

  1. Read seventh article about “hot new messaging app” that teens are flocking to
  2. Download the app
  3. Hem and haw over what to snap and who to send it to
  4. Delete the app

A couple of weeks pass.

  1. Overhear friend exclaim, “Did you see Matt’s snap with the cat? Soooo funny!”
  2. Re-install the app
  3. Wait around for someone to send me a snap. *Crickets*
  4. Delete the app

…this happened four or five times.

But then it finally stuck. Like white on rice.

A lot of smart people have talked at length about Snapchat’s secret sauce — making it fast and easy to create content, limiting that content’s shelf life, mimicking the personal and immersive quality of in-person conversations, feeding our curiosity for variable and unpredictable experiences, etc. I won’t rehash these. But I think two aspects of the experience have been overlooked.

Multiple ways to create content: Active messaging vs. passive publishing

The thing that finally made Snapchat stick for me was the release of their “Stories” feature. Up until that point, I was a passive consumer, basically only opening the app when I received a snap from someone. There was too much friction to publishing my own snaps — the issue usually wasn’t coming up with the content, but deciding who to send it to.

My Snapchat graph in the beginning was pretty small and composed mainly of friends I had met recently in Seattle; many of my closest friends weren’t using it yet. The personal nature of the medium can be a double-edged sword. On the one hand, there’s an immediacy and casualness to the interaction that makes it extremely engaging. On the other hand, it can inspire anxiety at the thought of imposing on someone who might not feel the same level of familiarity. The fact that you had to individually check each recipient’s name when sending snaps amplified this feeling of “I’m interrupting someone,” in addition to making it tedious to send snaps as your network grew.

But the ability to passively publish snaps and have them viewable to all your followers for 24 hours changed things. Suddenly I didn’t have to worry about the “who to send it to” part. I just posted to my Story and went on with my day. The key is that Snapchat closed the feedback loop by showing me the people who viewed my Story. Once I could see who was already looking at my stuff, it felt natural to send them a snap directly. Oh hey, a positive feedback loop!

Besides turning consumers into creators, “Stories” gave people another reason to open the app beyond receiving direct snaps, which meant more frequent and longer sessions. It also opened up a monetization channel for brands.

Matching emotional payoff with required investment

“It’s confusing.”

“How was I supposed to know how to do that?”

“I don’t get it.”

I hear this a lot from friends. Admittedly, Snapchat doesn’t have a great track record when it comes to straightforward UI, often burying new features in the settings, disabled by default. This would be a death knell for most products; people don’t have the time, patience, or attention to sit down and learn the intricacies of yet another messaging app. But Snapchat not only seems to get away with it, it actually thrives off of this dynamic. A few hypotheses as to why:

  • Snapchat’s seed audience, which evolved into its current base, is teens. This demographic likes to be the first to try new things and the density of their social circles ensures that interesting ideas spread quickly.
  • A novelty-loving user base is a big advantage, but it’s not enough. People will only invest in learning your UI to the extent there’s a proportionate emotional payoff to doing so. This is where Snapchat delivers — explorers unlock new filters, gigantic emojis, topical geotags, and other Easter eggs that make their snaps richer and more engaging. And as they discover and play around with these features, they spread it organically through their networks, piquing the curiosity of those they interact with.
  • Once people have sunk a bit of time and effort into your app, they feel more attached as a result of this investment. They’re more likely to tell their friends and take the time to bring them up the learning curve.


Tinder's secret sauce

Funny that out of all the dating apps I’ve tried—OkCupid, Coffee Meets Bagel, Grouper, HowAboutWe, Nerve, Hinge, Tinder—the latter’s the only one that’s resulted in dates. It’s also been the most successful in terms of number of matches and message to match ratio.

There tend to be two schools of thought when it comes to online dating: either optimize for quality by curating a smaller number of more targeted matches, or optimize for quantity and let people sift through the haul.

The problem with the first approach is that it assumes machine learning can predict what I want in a date based largely on self-reported characteristics. Sure, the technology will only improve, but it’s a tall order when I often don’t even know what I want for breakfast in the morning. Tinder’s design reflects the belief that dating is more of a serendipitous numbers game.

It’s fast

This is important when you’re dealing with a picture heavy experience. Hinge takes 5-6 seconds to load someone’s profile, which means it’s less likely I’m going to open it during one of those in-between moments when I’m in line or waiting for the elevator.

It’s super casual and engaging

Tinder positions itself as a spiffed up hot-or-not; not only does this keep expectations low, it offers a habit-forming reason to engage with the app outside of dating, which becomes a by-product. That’s why people “Tinder” with friends; as superficial as it is, it’s fun to make split-second judgments, especially in groups. The bigger point is that the more frequently you’re opening the app, the more opportunities there are for serendipitous encounters.

As a counterexample, Hinge only shows you matches who are second or third degree connections, which it surfaces very prominently. Because it relies on your Facebook graph, I’ve found this leads to one of two situations: your one mutual connection is someone you friended during freshman orientation and haven’t talked to since (more common), or you have 50+ mutual friends and significantly elevated expectations for the initial interaction.

Focus on the differentiators your customers care about

One of the unique opportunities that comes with working in a young industry is the chance to define what people should care about when they evaluate competitive offerings. Google did this in the early days of search when they trained users to value speed, breadth, and relevance, criteria they reinforced through visible cues in the product, e.g., ### results were returned in ### seconds. Of course, this only worked because there was significant overlap between what Google was good at and what users actually cared about.

Uber and Lyft are interesting case studies here given the different approaches they’ve taken to on-demand car service: Uber is more similar to a traditional cab company but with a professional bent; Lyft’s model is peer-to-peer ridesharing. In 2012 I was still using both apps pretty equally. Since then, Uber’s rising popularity seems to indicate it’s done a better job differentiating along the dimensions that people value in this space: price, convenience, reliability, and speed.

Lyft’s approach put it at a structural disadvantage in many ways. Ridesharing was a new idea and the company had to invest a lot of resources into educating customers about its benefits while allaying understandable fears about getting into a stranger’s car. The pink mustaches, fist bumps, and other community traditions went a long way and were a unique point of differentiation. But building trust was table stakes. And while Lyft was trying to convince people to just give it a shot, Uber was steadily improving on the factors that affect customer preference.

Beyond differences in approach, Uber has done a better job executing on discrete elements like price. Their marketing team has convinced me of the “fact” that uberX is always the cheapest option, regardless of whether it’s true. Lyft’s pricing is less transparent—in some cities it’s a suggested donation, in others it’s a set charge. Framing it as a suggested donation is an extra wrinkle. Even though most people probably never choose a different amount, introducing another decision adds friction to the process. Once you’ve experienced the freedom of stepping out of an Uber without having to think about paying at all, it’s hard to go back.


1. Spend one day assuming and acting as if every opinion you hear is true.

2. Spend the next day acting as if every opinion you hear is false.

Our natural inclinations fall somewhere between these two extremes; it can be useful to adopt a different default mode now and then.

Mess up early and on purpose

It’s easy to fall into the trap of being too precious about things you create, wanting to get every detail just right. This can be particularly paralyzing at the beginning of a creative process when there’s the biggest gap between where you are and where you want to be.

Over the weekend I was reminded of a good antidote.

A few friends and I spent Sunday working on several art installations for the group’s Burning Man camp. One of the tasks involved painting shuriken shaped stencils on pieces of what would eventually become a sedan chair.

I had only been painting a few minutes when Phil, a fifteen year Burning Man veteran and owner of the auto shop where we were working, strode over.

“I can tell this isn’t going to work.”

I hadn’t hit my stride just yet but the verdict seemed a bit premature.

“Can I see your brush?”

I obliged his command-disguised-as-a-question.

“What were you going to paint next?”

I pointed. Before I could pull my hand back, Phil streaked a big red line across it, dripping red splotches on the sideboard in the process.

“There. Now you can stop worrying about getting it on yourself or trying to be too neat about it.”

I was much more productive after that.


Talent signaling

An incredulous friend texted me the other day: “when are you guys going to stop putting new hires on the about page??”

She had a point. Navigating almost 600 facepiles is getting a bit unwieldy. I’ve already forsaken the scroll bar (comparable to traveling by Segway) and resorted to CTRL+F.

Interesting question, though.

When should a company stop listing all of their employees on /about? Why would you do this in the first place?

One hypothesis is that in the early stages, startups should pimp their employees—especially if they have A players—because it signals to potential candidates that they’re an attractive place to work with lots of other smart people.

After a certain point, however, the company grows big enough and has enough momentum to attract awesome talent regardless, and the risks of publicizing this information outweigh the benefits. Bios disappear when a company moves from poacher to poaching target.

Rethinking what it means to sell

Word association time:

I say “salesperson,” you say ________?


What’s ironic about these stereotypes is that they’re only true if you’re not very good at your job.

As I’ve shadowed sales calls over the past weeks, I’ve realized that the best aren’t “selling” at all in the traditional sense. In fact, I think the most successful salespeople actually view their job as two different roles:

  1. User researcher

  2. Problem solver  

The first requires the ability to listen, empathize, and draw out the underlying problem that you’re solving for. The second calls on your creativity, resourcefulness, and flexibility.

What this approach highlights is that the foundation of the entire process is understanding the use cases and paint points your product needs to address. Which means most of your time should be spent listening and asking thoughtful questions, not talking. I wouldn’t be surprised if there’s an inverse correlation between number of words spoken and deals closed.

This philosophy is helpful on a tactical level too. When you’re fielding questions or handling objections, try taking a step back to understand what’s motivating the concern. Otherwise you miss a valuable opportunity to gather context; it’s like being asked to pack a lunch without knowing if you’re going on a lazy picnic or an eight mile hike. To take a couple of examples:

Customer: “How much do you cost?”


You: “$795 a year for the first five licenses and $125 for each additional seat.”


You: “Annual billing is our most popular option and that’s $795 for the first five licenses and $125 for each additional seat. Are you looking to replace an on-prem server or switch from an existing cloud solution?”

Customer: “We have an on-premise server at the moment.”

You: “Ok, well let’s do a quick calculation of how much you can expect to save over five years…”

Customer: “I need something that’s more secure.”


You: “We encrypt your files at rest and in transit using the same technology as banks — 256-bit AES encryption and SSL for data transfers.”


You: “I’m happy to talk more about security. Can you tell me a bit more about your specific concerns and what your team needs?

Customer: “Well, we work with a lot of external vendors so the ability to set different access permissions is really important.”

You: “I see. There’s a couple of ways you can do that with Dropbox…”

How to disagree constructively

I’ve been in my fair share of meetings and have noticed that the best collaborators and leaders do one thing especially well: they know how to disagree constructively. That is, they can express a conflicting point of view in a way that moves the conversation forward and keeps the focus on the ideas instead of the people.

This skill also happens to be one that less effective communicators stumble over frequently, either by failing to voice their opinions assertively enough or bruising egos and their reputation along the way.

Liane Davey recently wrote an article in HBR with actionable advice for improving how you communicate a differing perspective. The key recommendations are:

1. Use “and” instead of “but.” It’s not necessary for others to be wrong for you to be right. Rather than try to override someone else’s idea, just add your reality. For example, “You think we should do X and I’m proposing Y. What are our options?” This engages your teammates in problem solving instead of competition.

2. Use hypotheticals to get people imagining a different scenario. This is especially effective if people don’t seem receptive to your ideas. For instance, “I hear your concern about XYZ. If we could address it, what could the campaign look like?”

3. Focus on the underlying goal. When you disagree with an idea, start by trying to understand what’s behind the suggestion. For example, “I’m surprised you suggested XYZ. What’s your aim in doing that?” Many conflicts arise over approach as opposed to the end goal. If you can agree that the problem you’re both trying to solve is the right one to prioritize, that’s common ground to start looking for a solution.

4. Ask about the impact. If you have concerns about a plan of action, ask people to think through the impact of implementing it. For instance, “How will this launch affect XYZ?” The key here is showing that you’re genuinely open to ideas and curious about the right approach.

5. Ask for help. If something really catches you off guard, be a bit self-deprecating. For example, “I think I’m missing something here. How will this help us with XYZ?” Be careful that your questioning doesn’t come off as insincere or leading. As with #3, people can quickly sense whether you’re making a real effort to understand someone’s reasoning.

360 feedback experiment

Inspired by this Quora post, a friend and I decided to reach out to people whose opinion we valued for feedback on our strengths and weaknesses. Here’s what we asked:

  1. What role does Alice play in your life? (How do you know her)
  2. What is Alice good at? Describe her at her best.
  3. What are some ways you’d love to see her build on these strengths?
  4. What is Alice not so good at? Describe her on an off day.
  5. What are some ways Alice can improve on her weaknesses?

What I learned

Three themes that came up most frequently in my off day feedback were: being more assertive, dealing directly with conflict, and handling unmet expectations. Alice and I actually shared a lot of the same areas for improvement and I was kind of disheartened to see how much of it broke down along gender stereotypes.

Most of these were things I knew in the back of my mind but hearing them crystallized by others makes me feel empowered to “get an ounce cockier,” as someone put it. It was interesting that more than one person pointed out I could benefit from behaving in what feels like an extreme way but that won’t actually be extreme at all.

Things that surprised me

Going through Alice’s feedback, I was surprised at how consistent the comments were after just a few (~3) responses. I imagine most of us think that only our family and closest friends see certain sides of our personality. The reality is that though the degree to which we express them might vary, our quirks and inclinations are relatively stable across contexts.

The majority of the feedback focused on the “how” rather than the “what”. Less of a surprise, but I was still struck by how strongly the comments skewed towards approach as opposed to accomplishments. How do you make people feel when you’re around them? What’s your process for tackling problems and disagreements? It got me thinking that there has to be a better way to incorporate this qualitative measure into the job application process.


Make sure your list of people includes a range of perspectives. Some dimensions to consider: relationship (friend, family, coworker, mentor, etc.), how long you’ve known the person, gender, and age (older points of view can be especially insightful).

Per the consistency point earlier, you really only need a handful of insightful responses to start seeing patterns. Alice and I had between 15 and 20 people on our list and a ~55% response rate. Some ways to improve this percentage:

  • Give people a heads up this is coming…so your dad doesn’t ask your friend if the experiment is real after the second reminder email (oops).

  • Insightful comments take some reflection. Allow two weeks for people to send in their thoughts.

  • We probably could have cut questions three and five as a lot of people ended up answering them in their responses to two and four. Additionally, while many of the suggestions raised were actionable, you’re going to be the best judge of what next steps make sense in your situation.

Use this as an opportunity to test your self-awareness. Write down what you think are your biggest strengths and weaknesses and see how it compares to others’ perceptions.

If you’re interested in giving this a try, here’s the email we sent and a spreadsheet to track responses. Would love to hear what you learn!

Dessert, video games, and excuse my language

On Thursday I attended a web development workshop put on by Girl Develop It. The audience, naturally, was all female. We started the session with a round of introductions where the last piece of information we were prompted to share about ourselves was our favorite dessert.

The next day, I went to a JavaScript workshop sponsored by Code Fellows. The event was open to everyone but the room was probably 80% male. When it was time to introduce the concept of JSON, one of the instructors launched into an extended Legend of Zelda metaphor. I realize that’s pretty mainstream as far as video gaming references go, but there was a level of understanding and satisfaction that escaped most of the women in the room.

Both Girl Develop It and Code Fellows are spearheading wonderful initiatives to promote programming literacy. My experiences this week just reminded me that gender imbalance in either direction comes at a cost. Sometimes the trade-off is worth it because it creates a welcoming environment for people to pick up valuable skills. And other times a lack of representation can lead half the population to think that programming isn’t for them. I hope some day in the near future I’ll attend a coding event where I’ll find points of connection with my peers and be challenged by a diversity of perspectives.

On a separate but related note, I’ve never understood that moment in a conversation when a guy turns to the women in a room and prefaces a curse word or other colorful phrase with “Excuse my language…” before breezing ahead. What’s the point? To preempt my ability to be offended? If you’re really worried about upsetting me, find another way to express yourself. Otherwise, just treat all of your listeners as equals. Not all men are crude conversationalists and not all women are sensitive ingénues.