1/ A few years ago a ton of “Uber for X” startups got funded, but very few of them - maybe none? - worked out. It sounds good but ultimately most failed on the supply side. Let’s explore why.
2/ Rideshare is special. Acquiring a broad base of labor for driving is expensive, often $300+. But then they can get requests all day. You can work 20 hours and even 50 hours a week if you want. You continually need the driver app to find new customers
3/ Where a lot of “Uber for x” companies fall down - valet parking, car washing, massages, etc - is that demand is often infrequent and there’s spikes at a few points in the day. What’s your supply side supposed to do the rest of the time?
There’s a TON of new startups coming from Uber alumni - I know of a half dozen in stealth, and Bird is already a breakout. It’s obv that the creativity and hustle required to make Uber work in its early years has trained hundreds of entrepreneurs. Very bullish on this group
As y’all know, Uber had a very decentralized mode of operation with each city being run as its own company. Each GM owned their P&L, hired their own people, and in the early years, would just put Facebook ads and other expenses on their credit cards! Great background
The product teams looked like this too. We had a “Programs and Platforms” model courtesy of Amazon / @jeffholden where each program was full stack, and the PMs ran hard again their mission/KPIs without introducing interdependencies
1/ When it comes to the future of media, there's a lot to learn from the Kardashians. Yes, the Kardashians. The future of content creators will look a lot like what they're doing today - whether you're a streamer, blogger, influencer.
2/ We've already seen the end result of mastering the ecosystem - $600M+ in revenue with only 12 full time employees. And tens of millions of followers. The full story here: forbes.com/sites/forbesdi…
3/ Today, a lot of creators are getting started on a single platform - be it YouTube, Insta, or Twitch - but as they pick up traction, they will increasingly be omni-channel. It lessens platform risk and allows them to build multiple touchpoints with users
One of the questions on scooters and ebikes has been- well what happened in China? Won't everything that happened there just repeat in the US? Piles of bikes, companies that have peaked out? The answer's no, because: 1) electric usage is different. 2) US rev/rider is much higher.
The current generation of marketplace startups has been incredibly successful. Airbnb, Lime, Uber, Lyft, Instacart, etc. I've been doing a broad survey of the best writing on this topic and wanted to share my list of 20 best links I've seen.
We look at a lot of marketplace startups at Andreessen Horowitz @a16z - and we fund a lot of them! - so it's great to compile all the best thinking.
@a16z 1. To lead off this list, my colleague @jeff_jordan has an awesome preso that covers everything from the marketplace "wheel" - network effects, and how they're different than ecommerce products. Amazing, thoughtful preso. Must watch.
1/ Building startups in San Francisco is getting ridiculously expensive. (No shit, sherlock). The NYT recently covered the phenomenon here: San Francisco Is So Expensive, You Can Make Six Figures and Still Be ‘Low Income’ nytimes.com/2018/06/30/us/…
2/ The major bottoms-up driver of this are rental/housing costs, which drive increased tech salaries. In turn, early startups are raising more and more funding, and paying their employees more. More funding in the startup ecosystem spins this cycle even faster.
1/ The Startup Brand Fallacy. Brand marketing is mostly useless for consumer startups. Startups build a great brand by being successful, finding product market fit and scaling traction, etc. But it’s not a real lever. Let’s not mix up correlation with causation!
2/ If this seems contrarian to you, it’s because there’s a vast ecosystem of consultants, agencies, and other middlemen who are highly incentivized to have you spend $ and effort on non-ROI/non-performant activities. Early startups should opt out of all of this
3/ It’s easy to confuse correlation and causation: If you’re starting a consumer startup, you see successful late stage cos with fawning media coverage, amazing conference speaking slots, celebrities on the cap table, etc., and think that’s what caused their success: Great brand.
1/ Startup advising is broken. We all know it. Entrepreneurs aren’t getting the value they want, nor are advisors. Instead, it turns into a series of dormant relationships where people are vesting way too much equity for no reason. We need to fix this!
2/ The typical relationship starts with an entrepreneur reaching out to a senior expert in the field. There’s enthusiasm to do something together! A few chats lead to paperwork, where the advisor earns as much as a senior software engineer over two years.
3/ Then, a few phone calls and a few busy months later, the relationship dies. The advisor is frustrated because there’s no owner for the relationship, and things aren’t going as well as they thought it would. If there’s an email, it’s to ask for help fundraising.
1/ "There’s nothing like some user testing to remind you that you don’t know anything." -@far33d, growth at @slack
2/ When you work on the same product every day for years, it's easy to assume that everyone uses the product the same as you do. Of course that's not true!
3/ IDEO has a wonderful phrase - "vuja de." Deja vu is when you see something for the first time, but think you've already seen it. "Vuja de" is when you see something every day but can see it fresh, like it's the first time
1/ Sometimes I regret that the term "growth hacker" got popularized. The content marketing industrial complex has turned it into a stream of tips and tricks -
"orange buttons = 50% lift!" - deemphasizing the old school wizardry implied by "hacker." Instead it's all hacks
2/ It's become an umbrella term for everyone that's working on everything from SEO optimization, email spam, viral content on YouTube/Instagram, to much more sophisticated systems-oriented approaches. As long as it's a quick hit.
3/ If you stop reading the listicles and quick case studies and start talking to folks on growth teams at the big tech companies in the Bay Area - you get a totally different picture.
1/ Random observation: One underappreciated advantage of companies that exist in the real world - Amazon Echo, Envoy, Limebike, Uber, etc - is the constant in-real-life reminders to try out and use the product
2/ The viral acquisition benefits are pretty obvious. If you've never seen/tried a product, but you see them swarming around your city (or your workplace), then naturally you'll want to try it out
3/ More importantly, some product usage patterns are naturally viral. Transportation fits into this bucket, which is why Uber's rider acquisition mostly viral/WOM
1/ I don’t trust hockey stick curves. Not on their own, anyway - which is often how they are presented in tweeted out infographics, demo days, etc. Hockey sticks can be gamed. They can show low-quality growth. And you need the right tools to figure out if they’ll sustain.
2/ The easiest way to show a hockey stick is to just buy ads. Buy a little bit one week, then a bit more the next, and so on. The curve will look good after a few months! Some of the acquisition will look attributable to ads, but it’ll cause some organic as well.
3/ That’s why you should always ask for a channel mix breakdown. How many of the signups are coming from which channels, and what’s the CAC and LTV from each. Sometimes they’re very different - 4x difference from blended versus per-channel.
1/ DAU/MAU has been a popular metric bc of Facebook, which popularized the metric. Theirs was always >50%. I found this from a Facebook 2004 media kit showing crazy high numbers even with a small base of 70k users: andrewchen.co/wp-content/upl…
2/ It's an important metric, to be sure, but it's often misused to say that "XYZ isn't working" when in fact, there's a slightly less frequent usage pattern that's still equally valuable.
3/ For consumer and bottoms up SaaS products, this metric is super useful, but seems to mostly exclude everything bseides messaging/social products that are daily use. These are valuable products, but not the only ones.
2/ So what do you do about it? One of the best case studies of this is from @drewhouston's Dropbox presentation from the early days. Lots of great stuff in this deck and it's worth paging through, now nearly 10 years later. Here it is: slideshare.net/gueste94e4c/dr…
3/ On slide 18, Drew talks about early experiments they did on paid search. They executed the industry best practices at the time - go to trial-based pricing, hide the free option, optimize landing pages. Slide: dropbox.com/s/7yvgxse97sxh…
1/ Have been writing nearly daily tweetstorms and am having a lot of fun doing it. They are much, much easier to write because you can just focus on the prose rather than going detailed on facts, figures, case studies, etc., as I usually do.
2/ Honestly it's been way more fun than I've thought it'd be, because it's easy! And the response from y'all has been strong. Thank you again for reading them!
3/ So far, I've written 5 over the last week. Here they are, if you've missed any of them!
1/ Many of the biggest implosions in recent history - especially ecommerce - have been due to startups getting addicted to paid marketing while fooling themselves on Customer Acqusition Costs. As spend scales, it always gets more expensive and harder to track - never less.
2/ A familiar story: New product launches. Nice spike, but it dies down. The product is low freq - gotta spend to grow. Marketing spend increases, it's profitable! More is spent, more money is raised via VCs. OMG this is working! Party!
3/ Suddenly top line hits a ceiling. Payback period goes from 9 months to 12, then more. Unit economic profitable, but not with staff + HQ. Without top line growth, more investment dollars can't be raised. Budgets get slashed, then layoffs.
1/ We all often talk about network effects in tech products like they're binary: Products are said to either "have" network effects, or they don't. I think we’re missing a lot of nuance that way.
2/ Let's take Hipchat/Slack's group chat as an example. When you add each new node to the network, the value increases dramatically! Up to a point. (~15 people?). After that, it gets noisy and the value increases more slowly / degrades for each new user.
3/ Uber is another example, albeit the two-sided version. When the network of riders/drivers are small, each new node decreased ETAs and increases utilization of drivers. Once ETAs are <5min though, the positive effect diminishes. It’s “good enough.”
1/ OK, this is an infuriating startup experience: You ship an experiment that's +10% in your conversion funnel. Then your revenue/installs/whatever goes up by +10% right? Wrong :( Turns out usually it goes up a little bit, or maybe not at all. Why is that?
Let's call this the "Conservation of Intent." For all your users coming in, only some of them are high-intent. It's hard to increase that intent just by making a couple steps easier - that'll just grow your low-intent users.
Doing tactical things like moving buttons above the fold, optimizing headlines, removing form fields - those are great, but the increases won't directly drop to your bottom line.
1/ Growing virally via mobile contacts and SMS doesn't work :( Big surprise for those of us who worked on email addressbook importing in the pre-mobile days, where it was key to build out 100s of millions of users for early Hi5, Tagged, Bebo, and eventually Facebook, Linkedin
2/ Email virality worked well for a few reasons: First, people actually liked getting invites! It was new and fun. Email clients didn't know how to filter user-to-user "bacon" so they let it through. But then the Law of Shitty Clickthroughs (andrewchen.co/the-law-of-shi…) kicked in.
3/ Second, email virality was often powered by a one-click "invite all" or "connect to call" kind of mechanic that might fire off 500+ emails in one go. This, combined with high response rates, was often enough to get viral factor >1. (This went away too, for obv reasons)
1/ Random thoughts on why I’m excited about the Work Graph - the network/interactions btwn people working together. Many diff cos tackling - Slack, Asana, Github, Invision, Dropbox, etc. - but it's just starting. The winners will be as big as the ones built on the Social Graph
2/ Startups building on the Work Graph have advantages: They can grow virally like consumer cos - with co-workers inviting each other - but they monetize like enterprise software. Their growth channels are rich and multi-faceted - ads/SEO/viral/etc work, but so do sales/PR/events
3/ Many consumer growth ideas haven’t crossed over yet. Email contacts importing drove consumer cos into billions of users. The tactics in the workplace should tap into calendars, who’s editing/comments on docs, team email aliases, org charts, etc. Who’s working on this?