@andrewchen

Get the newsletter · Featured · Recent

This is what free, ad-supported Uber rides might look like. Mockups, economics, and analysis.

intro

Cheaper and cheaper rides
Free, ad-supported Uber rides are inevitable, and if Uber doesn’t do them, a different competitor – perhaps Google! – will do it. It would be the next step in the industry’s trajectory towards lowering prices. Uber started in the high-end of the market, as “everyone’s private driver” which emphasized quality, but the early pricing was out of range for most. Dropping prices will increase demand, grow the market, and nothing will do that faster than going free.

Starting with UberX and now UberPool, each tier of service has dropped the price, tapping into more and more latent demand. Today, the figures are staggering – in San Francisco, Uber’s revenue is 3X bigger than the city’s entire taxi industry. Worldwide, the company now does 1 million rides per day and is adding 50,000 new drivers per month. Not bad at all.  I wrote a tweet about this concept a few weeks ago and wanted to expand more on this topic.

Free!
The next step is free. Lowering prices drive all sorts of goodness- more rides, busier drivers, shorter wait times – all which powers the company’s network effects. So how would you push the prices even lower than UberPool?

Well, here’s a crazy idea:

UberZero. A free, ad-supported tier of service that’s cheaper than UberPool and UberX.

That’s right, ads. I think it can be done well. And no, I don’t mean the cheesy in-taxi TVs that play ads nonstop, because those are broken for a variety of reasons I’ll cover. Instead, my proposal would be to put the ad units on your smartphone, in the dead time while waiting for your car and when you’re on your ride. And then to connect these ads with the big existing markets for app installs, lead generation, video ads, and local ads.

After all, advertisers are already comfortable paying a couple bucks for a variety of direct response actions from users. Individually, or together, you could imagine the above budgets making a big dent into the price of a single ride. In some cases, it might just be a discount of $1 or $2 off. In other cases, you might make the ride completely free.

Mockups for four potential ad units
Here’s four potential ad units that might work. These mockups were done in collaboration with the talented Chris Liu (@machinehuman), who’s done UX at Mercedes Benz, Alexa/A9, and IMVU. It was fun to throw together a couple designs.

Without further adieu, some of the concepts:

1) App Installs
The first and most obvious opportunity would be to tap into the ecosystem of paid app installs. Most of the $10B+ market for mobile ads is from driving app installs.

The user experience would be something like this- today, right after you book a ride, you often just sit there, staring at your phone, wondering how many minutes it takes for an Uber to travel a few blocks in San Francisco. Instead of just twiddling your thumbs waiting, how about getting a buck or two off your ride just from downloading and playing with an app?

Or similarly, imagine your current in-car experience. I often get into an Uber, pull out my phone, and do a little email during the ride. It would be easy to simply redirect my attention to an app, or a video ad, or some other activity that creates value for both myself and an advertiser.

It might look something like this:

app-download

As mentioned earlier, app developers are often paying $3-10 per install, especially in the games and commerce categories. This is a meaningful % when UberPool currently lets you go anywhere in San Francisco for $7 flat fee.

The biggest problem with this is that this is an incentivized install, which performs worse and Apple doesn’t like it. The plus side is that the folks who are really spending money in this area direct marketing-oriented and will back out the proper Cost-Per-Install to bid given all of that. As for Apple’s dislike of incentivized installs, you could certainly combine it with other ad units, and make the install a completely optional thing, which brings us to the next idea…

2) Video ads
One of the most compelling new ad formats for mobile is video, which are commanding high rates, especially when combined with a call to action to install an app. This is one of the many reasons why ad-supported companies like Twitter and Facebook are doubling down on in-stream videos.

So while you wait for your car to arrive, or once you get into the car, just watch a short clip that does a full-screen takeover of your phone. Then after you watch, the ad presents an optional action to download the app, which is a double dip on value. This packages both a brand ad with some a direct response ad unit, making for a powerful combination.

In Snapchat’s early tests on video advertising, Adweek and others have reported they are asking $750k, equivalent to a $100 CPM to participate. There’s no reason why Uber wouldn’t be able to command similar rates, especially for endemic advertisers in travel, shopping, and local.

It might look something like this:

movie-trailer

The key problem here would be to provide context and targeting. A classic TV ad against a show like, say, Oprah, provides both context, targeting, and a brand halo effect. An ad to an anonymous passenger isn’t compelling until you know a little bit more about them.

This begs the question, what do we know about a passenger? What would make sense from a targeting standpoint? Targeting is interesting because it’s an anonymous segment of people, which you could make off a number of interesting values:

  • Name
  • Contact info (email/phone)
  • Location
  • Where they’re coming from and where they’re going
  • Work or personal credit card

Of course from a targeting perspective, you’d never pass on the personally-identifiable information to an advertiser, unless there was opt-in (more on that later!). But how compelling would it be to for example target people in a segment like the following:

  • Passengers heading to a movie theater
  • Passengers coming from SFO airport, heading to SOMA
  • Passengers with email addresses like @google.com and @fb.com
  • Passengers traveling in a group of 2+ going out on a Friday night
  • Passengers heading home from work, or vice versa
  • Passengers heading to the Dreamforce conference
  • … and many more combinations

Now ask yourself, who else is able to deliver targeting ad inventory like this? There’s a very short list of companies that could even contemplate it. Of course it’ll be up to Uber as well as passengers to decide whether or not this kind of targeting criteria can be aggregated anonymously for ad targeting, but history has shown that ad targeting options tend to increase rather than decrease over time.

If you found the examples above compelling, imagine an instant rebate on your ride that relied on opt-in from the passenger to share this contact information directly. There’s a whole world of leadgen that’s willing to pay big bucks for accurate contact information for people who are in the market for targeted services and products.

3) Lead gen, especially for SaaS/B2B
I find leadgen an especially intriguing possibility because it’s commonplace for advertisers to pay something like $10 for an email address. For a SaaS/B2B advertiser, that number is even higher, as much as $30-50 in some cases. And if you’re talking about getting a credit card alongside an app install, that might be as high as $200 or more.

Of course you could never pass along a passenger’s contact info + credit card number without their consent. But what if you made it frictionless to consent? Then that previous list of attributes could all be sent in a series of confirmations: Name, contact info, payment information, etc.

At the heart of this is the fact that segmentation/targeting makes some audiences much more valuable than another. Today, a service like Uber charges a C-level executive the same price as the Average Joe. Imagine if you could have a leadgen campaign where an advertiser  decides that someone with a “name@google.com” address is pretty valuable given that they’re a Google employee. Or perhaps it’s interesting to know that they are coming from a Salesforce conference.

You might provide a call to action like this:

free-ride

This would make obvious sense for the high stakes world of SaaS/B2B leadgen, but it might also make sense to target a wide consumer base, especially around commerce. For example, targeting travelers who are arriving from the airport, to target them with highly personalized hotel/tour offers. Or targeting hardcore movie or concert-goers for their next night out.

Similar to the video ad scenario, the targeting capabilities would be key. Targeting based on location and context is key. And as I think about this, the most compelling part of the whole thing is that Uber/Lyft have data and context that is unique in the industry. They know where you are going, and where you are coming from, and enable you to go from Point A to Point B.

You know who else is like this and has created the most profitable and targeted advertising ever? Google, of course.

Google’s superpower is the Power of Redirection. When you search for something, Google can understand that intent and gently redirect you to the right place. Sometimes this is a destination that they deem relevant based on their algorithms, but sometimes, they just take you somewhere that’s paying them a bunch of money. This superpower is very valuable.

4) The Power of Redirection for local ads
Uber also has the redirection superpower, albeit in a different flavor. Similar to Google, when you plot a destination, there’s intent built into that. All the examples I’ve used previously indicate this- heading to Facebook’s HQ means something. Leaving from SFO means something. Heading to the 49ers stadium on a Sunday means something. And it would be easy for Uber to take you to your destination, but maybe they can suggest somewhere else to go, or perhaps they can suggest a complimentary product/service.

And using the analogy of organic versus paid search, perhaps there are organic destinations and there’s also paid destinations. In the case where a ride can be redirected to a paid destination, then Uber can literally, actually drive foot traffic. Pretty amazing!

For a trivial example, you might imagine something like this, that redirects you to a nearby Starbucks as a detour on your ride:

starbucks

CBXZ5V3W4AArWJn

(The second mockup that bundles in restaurant reservations was sent in by Marc Köhlbrugge, founder of BetaList – thanks!)

You could imagine a clear application for travelers of course. Heading to the W Hotel? Why don’t you try an Aloft hotel and get $X off your stay. Think about how compelling this is compared to other forms of getting local foot traffic. Billboards? Radio? Well, in this case you can literally get someone to your footstep.

I think this could be fundamentally a new market, although it would make sense to tie this back to the $4.5B mobile/local ad market. Or you could tie it back to the unit economics of something like pay-per-call, where local vendors are willing to buy a live phone call with a customer for $10. Nevertheless, still super compelling when you are talking about tens of billions of local ad dollars moving to mobile over the next decade.

Lots of practical challenges
But wait, could you really make all rides free all the time? Maybe not. After all, there’s aren’t infinite drivers. Won’t this make the ride UX horrible? Maybe, which is why you’d have to be thoughtful with the design. Another issue is that advertisers won’t pay for passengers to download the same apps over and over again – they’ll want frequency caps. And Apple doesn’t like incentivized installs, or maybe passing PII to advertisers will be frowned upon by the press. All very real issues that will have to be worked out over time.

And in the end, maybe it’ll turn into free* with an asterisk, limited to 1 per day within your city and only in the morning. Or just another way to decrease fares by another 25%. Whichever way it’s done, I’m convinced it will be done, because the benefits are too great.

Isn’t this perfect for Google?
If Uber doesn’t do this, perhaps Google will. There’s been rumors that they’re working on some kind of Uber-competitor, and they already have all the advertisers to make the above ad units work.

Providing free, ad-supported rides would certainly be a unique way to enter the market. Even more so, there’s network effects. Ad networks are fundamentally marketplaces, and they have tremendous network effects at scale. You could see the following virtuous cycle:

  • Free rides = way, way more rides
  • More rides = more advertising reach
  • More reach = more interesting targeting options, and drivers
  • More advertisers = more profitable yet free rides
  • Which leads to more drivers, customers, and more

At scale, a network like this would be very hard to attack, especially when combined with existing ad businesses like the type that Google possesses. Thus, if we see Google launch an Uber-competitor, I think it’d be a matter of time before they experimented with this.

Why old in-taxi advertising systems suck
The final point I want to make is to contrast all of these ideas with the old in-taxi ad systems we used to see. If you need your memory jogged, they used to look something like this:

IMG_5315

These systems suck. And they can’t generate the kind of economics to power the subsidy we’d want to see to make a dent on the actually fare price. The point isn’t to actually make money via advertising- the point is to drive down the cost of a ride so that the market is more efficient and liquid, creating network effects.

These old advertising displays are archaic:

  • Passive advertising content
  • Doesn’t know anything about you or your trip
  • Can’t get you to download an app
  • Limited internet connectivity

These systems are ineffective because ultimately, this display belongs to the taxi and not you. As a result, it lacks some of the key ingredients that an advertiser finds valuable: an easy way to get your contact info, or to get you to install an app on your phone. It’s good for display untargeted video ads and not much else.

That’s why Uber sits on a special opportunity to build an ad platform that’s never existed before. It’d be an ad platform that has unique access to context, intent, and built on your personal device. If it’s done well, advertisers will love it and consumers will be grateful to have free/discounted rides.

Yet another reason to be bullish about the company, in addition to all the great momentum they already have.

(Special thanks to Chris Liu collaborating on these great mockups that really make the discussion in this essay pop!)

PS. Get new updates/analysis on tech and startups

I write a high-quality, weekly newsletter covering what's happening in Silicon Valley, focused on startups, marketing, and mobile.

Personal update: I’ve moved to Oakland! Here’s why.

Oakland-JLS-2
My new neighborhood, in Jack London Square, Oakland, CA.

Where’d you move?
I moved to a tiny neighborhood called Jack London Square in Oakland. Yes, it’s named for that guy that wrote about the gold rush. I’ve only been here 3 months but I really like it so far. Previously, I lived in Palo Alto for 5 years, then about 2 years in the Lower Pac Heights neighborhood of San Francisco, but had never really spent much time in the East Bay. I had sort of heard that people were moving from SF to Oakland, but didn’t really have a reason to check out the neighborhood until a few people I know moved here.

Here were some of the articles I read while doing research:

PS. if you call Oakland “the next Brooklyn” to people who’ve lived here for a long time, they don’t like it :)

I live near there too! / How do I find out more about it?
If there’s interest, I’ll host a tech get-together or two.

Sign up here to get updates on an upcoming brunch/drinks/dimsum/whatever in Oakland.

There aren’t too many tech people here, so it’d be fun to get the small community that is out here together.

Where is it relative to San Francisco? How’s the commute?
I travel to the city pretty much every day. I usually take the BART, and sometimes the ferry (it has wifi!).

There’s a couple ways to get to the city:

  • BART (10min walk + 20min BART)
  • Ferry ride (25min ride + walk from Ferry building)
  • Car (30min without traffic, 60min+ with traffic)

I used to live in the Mission, and going from 24th+Mission to SOMA is about comparable to my current commute. However, I occasionally do have the morbid fear that there’ll be an earthquake while I’m underwater in the train.

Screen Shot 2015-02-24 at 2.07.42 PM

Where’s all the good food?
Right now, the Uptown neighborhood is opening the most new/amazing restaurants, where you can eat before you go to the Fox Theater or the Paramount for a show. Jack London Square has great food as well – there’s an eclectic mix of fancy pizza shops, vegan, and southern. A quick walk into Chinatown provides an endless supply of cheap eats, and the Oakland Chinatown is huge – about 2x the size of the city’s, without the tourist stuff.

A quick map search shows you where all the food is- pretty much in the Broadway/Telegraph corridor, but Rockridge, Temescal, and Grand Lake do well too.

Doesn’t Oakland have a ton of crime?
Crime was one of my top concerns moving to Oakland, but if the spectrum in San Francisco is aggressive/disturbed people in the Tenderloin to the nicest part of Presidio Heights, I think Oakland is about the same. You wouldn’t want to walk around Market St at 3am and you wouldn’t want to do that on Broadway in Oakland either. (Palo Alto / Menlo Park / Atherton are on a whole other planet, of course)

The biggest lesson I’ve learned from exploring the long list of East Bay neighborhoods is that Oakland is very diverse, and while the crime factor is a big one, it’s an acute problem for some neighborhoods and less of a problem for others. So, it all depends (just like SF, btw).

Houses in the Oakland Hills look like the kind of fancy houses you’d see while driving on 280 in the peninsula. Some neighborhoods like Rockridge, Grand Lake, and Adams Point are small and upscale, not unlike University Ave in Palo Alto. Uptown/Downtown feels like Market Street in San Francisco, but inexplicably cleaner. My new neighborhood, Jack London Square, feels a bit like South Beach in SOMA.

On the other hand, neighborhoods in deep East Oakland don’t feel very safe. That’s where you can find the car sideshows on YouTube.

Is it cheaper to live there?
For now, buying or renting seems to be about 50-75% the cost of San Francisco. Maybe as low as 30% if you are adventurous.

Is Oakland really warmer than San Francisco?
Yep. Sort of like the peninsula, up to 10 degrees warmer. Sometimes I miss the fog.

Here’s a typical day on the waterfront in Jack London Square.

Jack-London-Square_Bldg-F1-20-20-1200x556

Where do you go for coffee?
The headquarters for Blue Bottle Coffee is here. Yes, there’s usually a line.

But there’s also a bunch of other coffee places too:

Interestingly enough, the density of tech in Oakland is still relatively low. Cafes aren’t full of tech bros with terminal open. Coworking spaces are more likely to be nonprofits, writers, and sales, rather than unpronounceable names of startups. I’m sure a bit of this might change over time, and there’s been rumors of one of the big cos taking over the old Sears building in downtown.

bluebottlejacklondonoutside

How do I dress when I visit Oakland?
Like this video.

What’s the best way to visit Oakland?
The first step is to come out here by car/BART/ferry and check it out. I’d encourage you to do it, I think you’ll be surprised by how nice it is. And as I said above, if you’re interested in attending a casual get-together in the new neighborhood, or if you already live around here, just sign up on this mailing list and I’ll post some future updates.

PS. Get new updates/analysis on tech and startups

I write a high-quality, weekly newsletter covering what's happening in Silicon Valley, focused on startups, marketing, and mobile.

The most common mistake when forecasting growth for new products (and how to fix it)

weather
Forecasting weather is hard, and so is forecasting product growth.

Startups are about growth
Paul Graham’s essay in 2012 called “Startup = Growth” makes a big point in the first paragraph:

A startup is a company designed to grow fast. Being newly founded does not in itself make a company a startup. Nor is it necessary for a startup to work on technology, or take venture funding, or have some sort of “exit.” The only essential thing is growth. Everything else we associate with startups follows from growth.

The other important reason for new products to focus on growth is simple: You’re starting from zero. Without growth, you have nothing, and the status quo is death. Combine that with the fact that investors just want to see traction, and it’s even more important to get to interesting numbers. In fact, later in the essay, pg talks about how important it is to hit “5-7% per week.”

Getting to this number while trying to show a hockey stick leads to a bad forecast. Here’s why.

The bad forecast
The most common mistake I see in product growth forecasts looks something like this:

Image 2015-02-18 at 7.16.38 PM

In this example, the number of active users is a lagging indicator, and if you multiply this lagging indicator of a growth curve, it’s a truism that the growth will go up and to the right. If you do that, the whole thing is just a vanity exercise for how traction magically appears out of nowhere.

And of course these growth curves look the same: They all look like smooth, unadulterated hockey sticks. The problem is, it’s never that easy or smooth. In reality, you’re upgrading from one channel to another, and in the early days, you do PR but eventually that doesn’t scale. Then you’ll switch to a different channel, which takes some time but also eventually caps out. Eventually you’ll have to pick one of the very few growth models that scale to a massive level.

The point is, incrementing each month with a fixed percentage hides the details of the machinery required to generate the growth in the first place. This disconnects the actions required to be successful with the output of those actions. It disassociates the inputs from the outputs.

In other words, this type of forecast just isn’t very useful. Worse, it lulls you into a false sense of security, since “assume success” becomes the foundation of the whole model, when entrepreneurs should assume the opposite.

Create a better forecast by focusing on inputs, not outputs.

How to fix this forecast
A more complete model would start with a different foundation.

It would:

  • Focus on leading indicators that are specific to your product/business – not cookie cutter metrics like MAU, total registered, etc.
  • Start with inputs not lagging vanity metrics
  • It’d show a series of steps that show how these inputs result in outputs
  • And, how the inputs to the model would need to scale, in order to scale the output

In other words, rather than assuming a growth rate, the focus should be deriving the growth rate.

If you plan to 2X your revenue for your SaaS product, which is done by doubling the # of leads in your sales pipeline, and those leads come from content marketing – well, then I want to know how you’ll scale your content marketing. And how much content needs to be published, and whether that means new people have to be hired.

That also means that if you want to 2X your installs/day, and plan to do it with invites, I want to understand the plan to double your invites or their conversion rates.

Or better yet, say all of this in reverse, starting with the inputs and then resulting in the outputs.

Inputs are what you actually control
Focus on the inputs because that’s what you can actually control. The outputs are just what happens when everything happens according to plan.

One helpful part of this analysis is that it helps identify key bottlenecks. If your plan to generate 2x in revenue requires you to 5X sales team headcount when it’s been hard to find even one or two good people, you know it’s not realistic. If your SEO-driven leadgen model assumes that Google is going to index your fresh content faster and with higher rank than it’s ever done, then that’s a red flag.

In the end, it’s also true what they say:

No plan survives contact with the enemy.
smart prussian army guy

Keep that in mind while you fiddle around with Excel formulas, and you’ll be in good shape.

PS. Get new updates/analysis on tech and startups

I write a high-quality, weekly newsletter covering what's happening in Silicon Valley, focused on startups, marketing, and mobile.

The race for Apple Watch’s killer app

My top essays in 2014 about mobile, growth, and tech

Why messaging apps are so addictive (Guest Post)

IAC’s HowAboutWe co-founder: How to Avoid Delusional Thinking in Start-up Growth Strategy (Guest Post)

Mobile retention benchmarks for 2014 vs 2013 show a 50% drop in D1 retention (Guest post)

New data on push notifications show up to 40% CTRs, the best perform 4X better than the worst (Guest post)

Why Android desperately needs a billion dollar success story: The best new apps are all going iPhone-first

Early Traction: How to go from zero to 150,000 email subscribers (Guest Post)

New data shows up to 60% of users opt-out of push notifications (Guest Post)

Why aren’t App Constellations working? (Guest Post)

There’s only a few ways to scale user growth, and here’s the list

Lessons learned adding messaging to a notes app (Guest Post)

Retention is King (Guest Post)

Why consumer product metrics are all terrible

How to solve the cold-start problem for social products

How to design successful social products with 3 habit-forming feedback loops

Congrats to my sis Ada Chen, who’s joining SurveyMonkey as VP Marketing

How to make content creation easy: Short-form, ephemeral, mobile, and now, anonymous

My 2013 essays on mobile, startups, and tech

When a great product hits the funding crunch

A clever way to buy Facebook ads based on what your users like (Guest post)

Use this spreadsheet for churn, MRR, and cohort analysis (Guest Post)

Zero to Product/Market Fit (Presentation)

The Rise of Fat Venture Capital

How Google and Zynga set & achieve meaningful OKRs (Guest Post)

Case studies from “Why you can’t find a technical co-founder”

Easter Egg Marketing: How Snapchat, Apple, and Google Hook You

How is Yahoo really doing? Here’s the Google Trends data (Guest Post)

Ignore PR and buzz, use Google Trends to assess traction instead

Books I’m reading (2013)

Constrained media: How disappearing photos, 6 second videos, and 140 characters are conquering the world

The highest ROI way to increase signups: Make a minimal homepage (Guest Post)

9 ways a billion dollar new mobile company might be created (Guest Post)

Mobile traction is getting harder, not easier. Here’s why.

Why you can’t find a technical co-founder (Guest Post)

How to grow your app revenue with DuPont analysis (Guest post)

New college grads: Don’t sell your time for a living