Hey friends,
 
I just returned from a full week in Europe, visiting Amsterdam for business and Helsinki for the SEMrush Summer Jam. 
 
I’m pretty impressed by both cities. The last time I went to Amsterdam was probably 10 years ago and much has changed. I went to the TQ tech hub for the first time and met with some smart founders and marketers. But I also really liked how everybody spoke good English, was very friendly, and the city’s many little alleys with designer stores, bars, and restaurants.
 
SEMrush Summer Jam in Helsinki was a blast! Kudos to SEMrush for putting up such an event. I’m not sure I’ve been invited to something like this before: from a beautiful sailboat tour on the Gulf of Finland to bringing on Avinash Kaushik to speak about machine learning, data, and the future. I met so many smart people, had some inspiring conversations and interesting brainstorming sessions. I’m very excited to see SEMrush evolve over the next couple of months as a platform.
 
Cheers,
Kevin
 
  
Case study

Lime & Bird: pushing growth on the scooter market

The scooter craze is real.
 
Lime and Bird, two “electric vehicle sharing” companies (think: scooters for rent), are backed by the who’s who in venture capital. The hype around rental scooters is as big as the controversy. Lime is said to have secured a $250m funding round and is valued at $1bn. Bird has raised $300m at a $2bn valuation. Both companies have only been founded in 2017. Reason enough to dig into their Growth Strategies!
 
-read more-
 
First, the market. 500,000 vehicles squeeze themselves through San Francisco with a bit over 300,000 available parking lots. Not only does that leave ~200,000 vehicles without parking lot, it also makes the city unbearable in terms of traffic. Working and living in such a city means lots of movement and that trend is not shrinking. UBER and Lyft try to solve this problem, but there’s a big gap for the so-called “last mile mobility”: shorter distances that would take 10-20 min to walk*. Bird and Lime (and a couple of other, weird single name companies) want to fill that gap.
 
* To be fair, UBER and Lyft are getting in the scooter business, too. They see the competition on the horizon.
 
Second, the product. Rental scooters have lots of benefits:
  • Environmental friendly
  • No worries about parking
  • No traffic jams
  • Cheap ($1 per 30 minutes)
  • Fun
  • Easy to find
  • Simple to track
 
Third, Growth. Bird and Lime own the scooters, so their profit margin is around 10%. Customers can leave the scooters anywhere, but also make money by charging them at their own homes – smart! What looks like a hardware company is basically a two-sided market of riders/renters and chargers.
 
Lime grew to 150,000 users within 10 months! They have a really smart system to expand into other cities: letting users vote. Early expansion is tricky because you want to hit the right moment between first mover advantage and enough traction. If you scale too soon you could spread yourself too thin. Expand too late and it’s harder to get traction.
 
 
Getting a mobility product started demands huge loads of capital. Other than Lyft and UBER, who don’t own their cars (yet), scooters have to be bought and chargers have to be paid. 
 
The network effects are strong, which is always a good sign. I already mentioned the potential to build a two-sided marketplace. That’s network effect #1. 
 
The fun factor shouldn’t be underestimated. Seeing someone else riding a scooter while you’re sitting in traffic or walking by the street is very powerful. The product spreads through usage, which is almost like virality (#2). 
 
The third network effect stems from tracking scooters to find inventory bottlenecks, understand usage better and fine tune the product. Those are also referred to as data network effects.
 
Altogether, I think the scooter market could become huge. I’m also curious to see how that plays out in markets like Asia, where the car to bike ratio is very different from the USA.
 
Key lessons
  1. Use user feedback (votes) to decide when and where to expand.
  2. Look at the whole value chain to understand market disruption and future competition.
  3. Seek network effects at any cost, whether through data or building out market-sides.
 
– Comment

Avinash Kaushik on Machine Learning

As I wrote above, I attended SEMrush Summer Jam this weekend and they brought on a surprise speaker: Avinash Kaushik. Surprisingly, his topic wasn’t analytics (that came up a bit), but artificial intelligence and he had some provoking thoughts that I want to share with you, together with my own.
 
First, forget about demographics/psychographics and focus on intent. What people want to do in a specific moment is much more important than whether they’re 16 or 60, vote democrats or republican, or belong to Generation Z vs.Y.
 
Second, Avinash is not a fan of investing in channels that don’t allow you to measure and target behavior, like TV. He calls it “spray and pray marketing”. They don’t allow you to answer the question “what are people trying to achieve by following you?”.
 
Third, escape data puke. The future doesn’t belong to solutions that give you as much data as possible, but to solutions that give you the right data. Machine Learning will help us to analyze data and make decisions better than ever, so the challenge will be to understand which decisions are the most important ones. The shift is from data collection to recommendation.
 
Fourth, explicit programming will die out. To manually tell a machine what something is isn’t sustainable. Instead, we will need to provide programs the necessary criteria and limitations to learn by themselves.
 
Fifth, machines will be good a frequent high-volume tasks. Humans will tackle novel situations and use their creativity.
 
My personal thoughts revolve much more around how machine learning will transform our jobs. Simple tasks will be automated in the future:
  • Optimizing campaigns for PPC
  • Keyword research for SEO
  • Posting the right content at the right time in Social
  • Understanding which feature is used most often in product
  • Calculating customer acquisition cost, lifetime value, and other unit economics metrics
 
Instead, we should focus on asking better questions, find more creative ways to leverage data and look for new ways to teach and use algorithms. I think the implications for our careers could be huge and we should start thinking about it now.
 

Your weekly dose of awesome content

 
Bloomberg: Instagram is worth $100B
 
Search Engine Land: Google rebranded Adwords and Doubleclick to Google Ads
 
Notice how Youtube ads are highlighted. Google knows that Youtube is very powerful, similar to how Instagram is powerful to Facebook.
 
Sensor Tower: iOS 11’s App Store Increases Downloads of Featured Apps up to 800%
 
MediumHow we increased engagement using automatic personalized copy testing
 
Pinterest has productionized copy a/b testing.
 
Andrew ChenThe Startup Brand Fallacy: Why brand marketing is mostly useless for consumer startups
 
“Brand is a lagging indicator of success, not a growth channel.”
 
ChargebeeDear SaaS Peers, Scale Value, Not Usage (It’s Not as Simple as You Think)
 
I like the perspective of pricing according to core product value.
 
First Round: Here’s What a Real Growth Strategy Looks Like — Road Tested by Facebook and Remind
 
TechCrunch: A leaked look at Facebook’s search engine for influencer marketing
 
Bonus