Why I want to build Lyft for Kazakhstan and why it can fail

Update: Having torn a page from the lean playbook, we just launched [-1] an MVP for the taxi service described. No surprise here — WE are the drivers!

Part I: Peace

On January 14, I came back to my hometown of Astana, Kazakhstan after an awesome month long vacation in the US. One service I unexpectedly fell in love with while there was Lyft. After 4 days in New York City and 4 days in San Francisco Bay Area, we arrived in San Diego International Airport where we got picked up by our Airbnb host. She mentioned coupons from Sidecar and Lyft — a little unassuming event which led me down the cab catching ‘rabbit hole’ as the last time I used Uber was a year ago — back in December 2012. The distributed city of San Diego is very conducive for taking cabs. I gave everyone an equal chance, from Christmas to New Year’s Eve having cumulatively spent $133 (7 rides, 53 miles, 115 minutes) on UberX, 43 bucks on two rides on Sidecar and $198 (13 rides, 81 miles, 181 minutes) on Lyft. Well, almost equal — Sidecar and a few cars it had in San Diego didn’t hold up to the fierce competition. Lyft won my heart over and the next year — from January 7 [0] to January 12 (the day we flew out from SFO), I spent $243 (10 rides, 77 miles, 178 minutes) solely on Lyft rides [1].

Hmm, I wanted to write a pure essay when I just started typing this, but looks like out of laziness and respect to my mechanic memory [2] I’ll resort to the lowest common denominator of all — I’ll give you the list of k + 1 reasons why I want to build Lyft experience for my hometown [2.5]:

• Having done quite a few pure online or software projects, I really want to do something involving the real world. The intersection of technology and the offline physical world is really appealing. There are lots of people out there who don’t spend all their time in front of the screen or their iOS devices. It’s easy to miss, but the physical world is really high bandwidth — and lots of things can be made here.
• You can say, ‘What about innovation? What about starting from scratch and building something which never existed before?’ I always looked down on copying, but now I realize that there is something there. Especially when you’re just starting. ‘Start copying what you love. Copy copy copy copy. At the end of the copy you will find yourself’, as the notable Japanese fashion designer Yohji Yamamoto once commented. It’s better to steal and build a copy than to always dream of making something truly original and in the end not building anything at all. To innovate = to copy + to improve. An attentive reader will notice that I just copied Evernote’s Phil Libin in the last sentence. And I’m not talking about uncreative and shameless plagiarism, I’m talking about taking something great, taking it apart and seeing how it works. Having said that, the reason I want to build on-demand car sharing in KZ and in Almaty in particular is not because I ran a Markov process on the corpus of Hacker News articles to mine good startup ideas. If I were to look at the hot companies from Crunchbase, what prevents me from saying, ‘I want to build Netflix / Lending Club / Airbnb / Kickstarter / Square / LegalZoom / Zocdoc / etc. for KZ’. In fact, the equivalents of none of the services I just mentioned are present in Kazakhstan at the moment. So if you start companies not based on your personal problems, talents or passions, but based on the intent to clone and localize the successful western models, go for it! There’s no shortage of things to work on!
• You can say, ‘In Almaty it’s much faster to stick your arm out and just get picked by a random taxi driver. Taking the phone from the pocket, launching an app, pressing the button and waiting is just too much’ [3.5]. That’s true, and we’re not planning to fight the natural tendencies. However, there are around 40 traditional taxi dispatch services in Almaty with the loyal customer base — those who due to circumstances or personalities can’t afford to play the ‘Russian roulette’ on the streets hoping that the passing car is heading the same direction or the driver is in the mood for spare cash.
• If you forget (or remember) about the possibility of getting picked by unofficial taxis, the current state of taxi and transportation in general in Almaty, Kazakhstan (where I’m currently residing) is miserable, to say the least. You’ve got to google for one of the dispatchers’ phone numbers, call it, wait for 30 minutes, wait again, call to the driver a few times to explain that you’re on other side of the street, and wait again. The alternative is using an iPhone app where you have to go through 7 screens and answer 5 questions [3]. I just want to push a single button and the car arrives out of the blue. That’s on the demand side [3.5]. Things are not so much better on the supply either. You are either getting a fixed salary working crazy hours or operate under the constant pressure to meet the daily or monthly minimum. The wind of change is blowing in our direction!
• The service industry in the former Soviet Union republics, to put it mildly, has a lot of room for improvement. From the gloomy faces welcoming you at the passport control at the airport, to the rude waiters/waitresses and very unfriendly staff at the department stores who don’t even bother to answer simple questions like, ‘Do you have the M-sized sweater?’. Everyone sees this and takes it for granted. ‘That’s the way things are. Don’t bother changing anything’, they like to say. Most of us complain and move on with our busy lives. The best way to complain about the service is to provide a better one, I think. Just bring a candle into the dark room instead of fighting with the darkness directly, the darkness will fade away. You’ve got to start somewhere. Taxi seems a good place to start. I use it. Lots of people with or without cars need it in the city.
• I was absolutely amazed by the airplane experience provided by the Lyft drivers: complementary bottles of cold water, gum or chocolate. You can charge your iPhone or ask them to play your favorite playlist. It doesn’t end there though. I heard one driver in LA was telling me how his acquaintance offers her passengers to sing karaoke at the back seat. Another driver bakes two batches of cookies every night and offers each flavor to his passengers. They taste each cookie and vote for their favorite. The cookie, winner by the majority vote at the end of the day, will battle with a new recipe the next day. Lyft doesn’t pay for this obviously, drivers do it for their and passenger’s enjoyment. Some drivers wear Batman costumes, do the hip hop or Harry Potter trivia for their passengers. Rides become theme parks to awaken your inner child.
• One of the questions potential Lyft drivers need to answer at the interview is ‘How will you make your Lyft experience special?’ And every Lyft driver I had a chance to meet seems to be doing their best at being unique. Lyft for drivers becomes not just their source of income enough to quit their boring consulting jobs. It’s the vehicle for self-expression (pun intended). You can finally have fun at your job and free your quirky personality from the cage of conformism and conditioning. You can pursue your hobby while at work. If I were a Lyft driver [4], I would totally wear my yellow jumpsuit from ’The Game of Death’, quiz my passengers on Bruce Lee trivia and let them enjoy ‘The Way of the Dragon’ at the seat in the back. And as for many fellow Lyft drivers, this won’t be done for extra tips or other material nonsense, I’ll do it because it’s fun and because that’s how I feel and express myself here and now.
• It’s not just a transportation service, it’s about building a community of friends in a big city. It’s about socializing and sharing your positive energy. Once we graduate from college, our circle of friends usually consists of those we met there or at the office. Spending 9 hours of your daylight in the realms of the office substantially limits exposure for new encounters. Imagine that a simple ride to the Nippon Kenpo school you practice at in the evening can offer you something more than just a reliable way to get there safely and in time. Hell, the driver him- or herself can be an experienced martial artist! The Lyft drivers I met ranged from rock band musicians to consultants in the movie business and airport personnel.
• The side effect of providing a ride is that you could exchange business cards with your passengers, connect on Facebook, or even promote your small business. A few musicians who gave me lifts in the Bay Area distributed the brochures and shared info about their upcoming concerts.
• I don’t want to sound like a technologist who selfishly increases the technical difficulty (having forgotten one of the key principles of Apple philosophy famously stated by Steve Jobs at WWDC 1997 [4.5]) and gets stuck in the ‘one year, 0 releases’ perfectionist’s trap, but the effective transportation solution offers lots of great computer science problems. To keep it short, you’re minimizing the wait time on the demand side and maximizing the utilization on the supply side.
• One of the billboards on the Oakland bridge I saw had ‘Work 9-to-5’ crossed out and ‘Become a Lyft driver’ sign written underneath it. Quite a few Lyft drivers decide not to keep their day jobs and work flexible hours driving their pink mustache cars.
• There are generally 2 types of people: thing people and people people. “Thing people enjoy hearing about new technology and other clever tools and possessions. They also enjoy discussions of processes and systems, including politics. People people enjoy only conversations that involve humans doing interesting things. They get bored in a second when the conversation turns to things” [5]. Using this vocabulary, I’m more of a people person — I love seeing that I made someone (not something) faster, higher, stronger, happier, better or cleverer. Those things excite me. Building Lyft seems to be a great fit. You’re connecting people in the real world, it’s the people business. [6]
• Since the price is determined by Lyft, it frees you from the burden of negotiating, or deciding what tip is appropriate — saving you from decision fatigue. As the money is out of the equation, just have a seat and enjoy the ride. Win-win for both sides!

You’ve just completed the idealistic ‘Peace’ part of my dilogy. Scroll down to continue with the dramatic ‘War’ piece.

Part II: War

Start with the end in mind. The valuable exercise is to think of as many reasons why we could fail as possible. What are the potential reasons Lyft model might not work in Kazakhstan?

• It might be hard to find drivers who are social and emotionally intelligent. From now on, it’s the driver who should take the responsibility to break the wall between the customers and the provider. The current state of things is that you’re ‘welcomed’ by an indifferent and quite exhausted taxi driver. All he cares about is 500 tenge [6.5] he gets at the end of the ride. Greeting your customer and a fellow Lyfter and genuinely asking ‘How is your day going so far?’ can’t be a lot to ask [6.6]. The series of interviews will ensure that every driver passes the bar. We’ll have to do special training sessions if necessary. Rewiring the old Soviet operating system is a big undertaking.
• The established belief is that the drivers in Almaty/Astana are uneducated bunch who came from the countryside and have to drive taxis as they can’t find work elsewhere. It’s time to work from first principles and question whether this is the case. Is it possible to attract creative professionals, freelancers and students [7] by providing flexible schedule, competitive compensation and awesome passengers?
• We might have problems recruiting the drivers and too few drivers to fulfill the demand [8] can lead to considerable wait-times for passengers. And a huge temptation to just raise a hand (or a leg if you remember that scene from ‘Enter the Dragon’ where O’Hara breaks into Lee’s room while he’s practicing) and get a lift from a passing car. As far as I know, Lyft launched publicly with around 150 drivers in San Francisco. Almaty has the land area of roughly thrice as SF, the population of Almaty is around 1.5 million vs 825,000 in SF. Expecting higher (but not by much) smartphone penetration in SF, I wonder what the right number of drivers for public launch in Almaty is?
• We might pick the wrong platform to build the app first. Supporting both leading mobile operating systems from the get-go is a suicide in the light of limited technical resources. In the US despite the shrinking share of iOS devices, it’s no-brainer: iOS first, Android second. Based on the official number of smartphones sold in Kazakhstan last year [9], iOS has only 7% and Android — around 80%. However, the ratio is more like 1/3 to 2/3 based on the usage data from a few mobile applications my friends are involved with. It will be great to see the stats from the key local mobile operators (Kcell and Beeline) to get a much better estimate, but those numbers aren’t public, as far as I know. If I were to build it for myself, I would go iOS first. Never owned an Android in my life [10]. On one hand (demand), the majority of the people [11] I’m surrounded by here are iPhone users. They tend to have this early adopter mentality, as the folklore has it. And we need it to cross the chasm. On the other hand (supply), almost all the drivers I see now own Android smartphones. The third alternative is obviously building the web app using technologies like Meteor and PhoneGap [12] and deploying on mobile via the web view.
• My hypothesis that we can attract drivers whose primary motivation is to meet interesting people, not money needs to be tested asap. We don’t want to go into dating, mind you. The aim is to provide an experience as if you’re picked up by a friend. Building the offline community with the help of technology — that’s what drives me.
• The elephant in the room is safety and all things legal. Being picked by a complete stranger is nothing new for Almaty citizens. But background checks, interviews, car inspections, criminal and DMV record checks is an absolute must. The legal issues create a huge risk, but it’s not something I’ve dove into yet. We might be seriously upsetting the existing market players and city officials.
• Another missing component is a good payment system. The transaction won’t be cashless in 90% of the cases. Locals are still not quite comfortable using the debit/credit cards, plus most of the cards issued by local banks aren’t ready to be used online. At least without a phone call to the bank representative with a request to ‘unlock the card for a month or two’. Instead of complaining, my proposition is the following. The passengers will pay using their most comfortable option — be it cash or a card (they need to add to the system only once). The 20% commission from the drivers will be taken from the Qiwi-wallet, a popular payment system with 32,000 terminals across the country.
• Our citizens are not used to metered cabs and accustomed to knowing the fare before the ride. What happens if they don’t know the price in advance? We should definitely provide them with an estimated cost and show that’s either cheaper or on par with the traditional taxi dispatchers. I’m not planning to compete with the unlicensed taxis on price. At least for now.
• Two-sided market like this comes with the natural and tricky chicken-and-egg problem. For ride seekers to come, the sufficient number of drivers should be in the market. For drivers to come, enough people should be looking for rides. We’ll focus on building the supply first by heavy recruiting and doing the thorough personal interviews with the drivers. Quality (drivers) > quantity (drivers).
• Another reason is that my plan for the private beta launch might fail. Currently it entails building an invite-only system to test the demand and offering rides in a limited area in the central part of the city during the commute hours (7:30-10am, 5-7pm).

Building Zappos of on-demand car sharing in the city where more than half of all the car seats are empty is exciting. Let’s see what the future holds for us: will this essay be an example of ‘The Blog Minimum Viable Product’ or just ‘words, words, words’ [13]?

I have two questions before we go. 1) What are the obstacles for smartphone powered peer-to-peer ridesharing in Kazakhstan I have overlooked? 2) iPhone or Android? Appreciate your comments.

Notes
[-1] ‘If you’re not embarrassed by your first version, you launched too late’ -Reid Hoffman.
[0] We rented a car, so didn’t need one until January 7.
[1] Take a look at the usage patterns on the basis of 200+ Lyft rides from my friend George Xing.
[2] And to save everyone’s time after all!
[2.5] I’m not talking about the entire country yet. As Paul Graham noted in one of his essays, ‘The best way to do big things is to start from small things and doing them well’ (don’t remember the quote verbatim).
[3] Exaggerating to my advantage here obviously.
[3.5] The most popular way to catch a cab in Almaty (just like in Moscow and across pretty much the entire former Soviet Union) is to stick an arm out and let an unlicensed stranger drive you where you need. ‘Everyone with a car is a potential taxi driver in Moscow’, as they say. I’m not planning to compete with that. Human inertia is a difficult beast to beat. Our only competition in the beginning are the numerous traditional dispatcher services. We’ll be competitive price-wise, but beat the hell out of them on quality.
[4] First things first. Gotta learn how to drive first!
[4.5] Steve Jobs: “You have to start with the customer experience and work backwards to the technology” [video].
[5] Great observation by Scott Adams in his ‘How to Fail at Almost Everything and Still Win Big’.
[6] It’s important to note that our primary focus is still more efficient and affordable transportation. None of the complementary things (great conversation or a free refreshing bottle of water) will matter if you, as the passenger, are late for your meeting.
[6.5] The local currency of Kazakhstan
[6.6] Obviously, if the customer is not in the mood to talk, he just tells this to the driver. Seriously though, overly talkative drivers is not the first thing I’m afraid of.
[7] With around 50 institutions of higher education, Almaty is clearly the Boston of Kazakhstan.
[8] Too many ride requests? What a ‘great problem to have’! I smell an optimist once again.
[9] #1 & #2
[10] Nexus 5 looks really cool though.
[11] 90% — if I were to guess.
[12] Thanks to my friend Slava Kim for the suggestion.
[13] William Shakespeare, Hamlet.

Say hi @suleimenov