Follow by Email

Search This Blog

Jun 28, 2017

The bike sharing hype - Overview

13 of the world's 15 biggest bike sharing fleets are in China, but the two others are London and Paris, showing the potential for Western megacities, too. This lengthy article in the Guardian gives a great overview of pros & cons and provides some stunning numbers on how fast the "Uber for bikes" companies have grown, and what effect this can have on mobility in urban areas (spoiler: there are serious cons as maintaining the bikes with human labour might be more costly than just adding new ones to the fleets):

Autonomous cars to capture 4 terabytes of data every 8 hours of driving

"Big" data is selling it short. 4 TB of data for 8 hours of driving - given that autonomous cars will likely be driving most of their time instead of sitting there as dead material on some parking lot - is unimaginable. I never saw it from this perspective - and I don't know how much of this data can be processed via the onboard network, where performance can probably be solved easily - but if a good portion of this data has to make its way to the cloud and acted upon quickly, we might have a performance obstacle to overcome before the takeover of autonomous cars can take place.

Uber should test a robot CEO

Interesting thought by Nicholas Carr: A company that consists of software and data as its main assets, and that just lost its CEO due to human flaws, not managerial shortcomings, may be the best case to test how to "automate the automators". Probably this would fall under the categories of deep learning and narrow AI which, as far as i know, we have pretty well under control. In fact, Carr says "a two day hackathon" would probably be enough to create a CEO. Good, quick read:

Jun 9, 2017

Commerce in Messenger / GIF Demo

Facebook messenger is very likely to become the next big platform in a couple of years. It has the potential to combine our two main ways of organizing and accessing information: intentional use and discovery. Both are also major mechanics in online sales. Here's a demo of how a sunglasses purchase could happen on messenger, utilizing the whole infrastructure: 

Amazon's dominance in one chart

Click it. It says it all.

What Bleacher Report learned from Snapchat Discover

The clear leader in social media sports journalism, Bleacher Report, has 11 people working on Snapchat Discover. That is huge and i cannot imagine this activity being profitable yet. Unfortunately the article does not reveal numbers, but for anyone thinking they could manage a Snapchat account "on the side", here are some learnings that may convince you otherwise:

Jun 8, 2017

Understanding Blockchain

I'll admit it: I am far away from understanding blockchain. But I am reading up on it and so far, it didn't prove to be of immediate necessity to know this for my business, so I am taking my time. If you are in a similar position, here's one article that adds a new perspective and is accessible to me, TechCrunch:

Intro to cryptocurrencies, or what does Ethereum at 260 USD mean?

Crazy times. A few weeks ago, Vinay Gupta, the release coordinator for Ethereum and a prolific blogger, consultant, publicist wrote a piece about what it meant that Ethereum hit 100 USD. Two weeks leater he updated it, because they broke the 200 USD mark. Now that I am writing this, again a few weeks later, we're at 260 USD. But the piece is not about the next bubble we're entering or how you and I are total idiots for not fully understanding what is gooing on; instead, it is one of the few articles that actually helped me understand that whole concept. If you, like me, never met a person who never stops talking bitcoins, smart contracts and blockchains, but who also could really enlighten you about these topics and could even answer some of your (maybe dumb) questions to some degree, this article is worth your time.

Why measuring customer preferences is an illusion

We all know the quote with the faster horses, even though Henry Ford probably never said it. But it is very useful in order to understand that innovation almost never comes from what consumers tell you in a survey. There are reasons for that, and similar reasons apply to customer data in general. This article points out why customer data from studies or research might be helpful in many ways, but can't deliver a blueprint of what to build next (and please note, this does not refer to data constantly generated and analyzed through product usage, but to the "traditional" survey data many use). Interesting read: