Follow by Email

Search This Blog

Oct 13, 2016

Chatbots overview

I am still waiting for that breakthrough chatbot on Messenger, but everyone I talk to is excited about bots - so am I, but I believe we are still in the stone age of bots, even though I have worked on their then-equivalent some 15 years ago. There's enough reason to be optimistic about the future of chatbots, and here is a good overview on the bot environment and list of links to useful resources at the end of the article:

[German] Auslandswachstum der deutschen Fußball Bundesliga-Vereine auf Facebook

Sehr nett aufbereiteter Überblick über das Fan-Wachstum der Fußball-Bundesliga-Clubs auf Facebook - aus dem Ausland.  Der Anteil deutscher Fans auf den Facebook-Seiten von Bayern München liegt unter 7%, bei Dortmund unter 12%, Tendenz sinkend - im Hinblick auf Auslandsvermarktung aber eine gute Nachricht. Bei Klubs wie Leverkusen oder Bremen treiben Spieler (Chicharito, Pizarro) Horden von Fans aus bestimmten Märkten auf deren Seiten, aber man sieht wenig Anstrengungen seitens Club oder Liga, diese dann in allgemeine Bundesliga-Fans zu konvertieren, wie die entsprechenden Spieler weiterziehen. Und für das Auslandsgeschäft wird es wichtig sein, neben der in der Champions League vertretenen Spitze Vereine wie Köln oder HSV im Ausland mit Profil und Anhängerschaft zu versehen.
Schöne Zahlensammlung, leider wenig Interpretation hier:

Importance of software for automotive companies

Car makers are good at building tradtional, "machine engineering" software. Software that optimizes how the tires behave while driving a curve, while breaking, or software that checks the status of the engine or any other functional part of the car. When it comes to consumers, car software always sucked. The "onboard computer", it feels, has barely left the C64 stages only in recent years. It is still way easier to bring my music on my phone and then connect the phone via bluetooth to my car speakers, than - what an innovation - just synchronize my music collection with some hard drive in my connected car, or, imagine that - stream from Spotify or Soundcloud or anywhere right in my car. And why don't I have a Amazon Echo/Alexa or at least Siri Assistant in my car for years? And despite all the progress made by car companies, collecting, analyzing and acting upon data, ideally in real time - compare Waze for example - is a whole new game. Hiring? Takes too long. Acquiring? Yes. Integrating? Pray and hope for the best. Here's an interesting overview from a venture capital perspective, on deals made in the past and companies that may succeed in this area in future:

Publishers on Instagram & Snapchat stories

It is hard to convince companies and especially publishers to invest in their Instagram and Snapchat teams as monetization or traffic generation is far away from the levels Facebook can provide; therefore, many agree to "play around" on these channels, but are reluctant to really invest in resources and learnings without getting an immediate return. Seen from Germany/Europe, the situation in the US seems to be a bit different. Here are two interesting links:

How Snapchat is used by publishers, with statements from HuffPo, Mashable, NPR, TheVerge and others:

And here's why Business Insider favours Instagram stories over Snapchat:

Online vs. TV: Presidential debates

We always use highlights - Superbowl, breaking news, or presidential elections - to measure the state of digital in comparison to other, traditional media channels. I believe we need to careful with these, but they are still great indicators about how (audiovisual) content is distributed. This article shows a nice overview of numbers for the second presidential debates. While TV reported a viewership of 63 million (let's ignore for a moment how they measure it, and how many unique viewers these were, or how many minutes were watched per unique vier per average), the Facebook live stream reached 7.4mn viewers. And YouTube's videos round the debate aggregated some 124 million views (although here too, we don't know the uniques). While TV viewership from debate I to debate II declined 20 percent, the YT views went up 40%. Maybe that's an indicator for a viewing behaviour of people who, if they don't follow the full event on TV (or something else on TV), using YT and other digital channels to catch up and watch highlights. Overall, YT views for presidential debates grew sixfold compared to the 2012 views. More numbers here:

The problem with open floor plans

I haven't regularly worked in an office for more than 10 years now, and of the 10 years before, I maybe spent 2 or 3 years on an open floor. For the rest of the time I had the luxury of an own office. When I visit clients, I realized that in recent years, even high executives moved to open floors. It is supposed to be a more transparent, open atmosphere, a more communicative environment, everyone is accessible. If you ask me, for someone who tries to get things done (other than emails or signing invoices), this doesn't really help productivity. Although putting together strategy presentations in PPT and coding may be two very different things, I can at least relate to the programmers being tired of constant interruptions in their work. Studies have shown that a programmer needs 10 to 15 minutes to get back to being productive after an interruption, and that on average, they only get one 2-hour-uninterrupted-session per day. God save the home office.

The importance of context for innovation

Great interview with the author of one of my favourite books, "The Innovator's Dilemma", Clayton Christensen. The book is from 1997 and can still be adapted to today's business world (show me another book about innovation that does that). Christensen talks about necessary additions to his theory of disruption, mostly social and emotional dimensions of products/services that may have more impact than we thought so far. Interesting red. 

Oct 3, 2016

[German] Traffic Quellen

Die Zeit hat in der Auseinandersetzung mit dem diabolisch genialen Digital-Guru Günther "Neo" Oettinger ihre Trafficquellen veröffentlicht. Interessant dabei, dass Search&Social weniger als 25% des Traffics ausmachen (Facebook nur 10%, Google nur 22,5%), und wenn man Dark Social radikal rechnet, auch nur knapp an 40% kommt. Direct Traffic: über 57%. Das kann man jetzt so lesen, dass die Besucher dort im Schnitt halt 109 Jahre alt sind und ein Internet von 2006 verwenden, oder dass es in den Bereichen Search und Social noch enorme Wachstumspotenziale gibt. Spannend auch, dass viele (wie ich) sich über den hohen Anteil Direct Traffic wundern - das ist sicher nicht mehr der Normalfall. Interessant ist es allemal:

Pokemon Go reached 500mn revenue faster than any other game

Winter is coming up, we want be outdoors that much, Pokemon fatigue is setting in and I see all these people arguing that PokemonGo was merely a "one hit wonder". If so, i'd like to have one too, please: No other game app ever reached 500mn revenue that fast. In fact, it reached that milesstone more than 3 times faster than Candy Crush Saga or more than 6 times faster than Clash of Clans. So no matter how the winter goes, we can't say that thing was a flop, can we?

Facebook Newsfeed algorithm overview

Good overview of what we currently know about Facebook's newsfeed by Techcrunch. I guess it still comes down to engagement, engagement and engagement as the three main criteria to expand organic reach, but it is interesting to see how this thing evolves over time.

[PDF] Stanford study on artificial intelligence

The report is named "artificial intelligence and life in 2030", and of course, since these areas grow exponentially, predictions are hard to make. Plus it is a long read. The Exec summary closes with:

"If society approaches these technologies primarily with fear and suspicion, missteps that slow AI’s development or drive it underground will result, impeding important work on ensuring the safety and reliability of AI technologies. On the other hand, if society approaches AI with a more open mind, the technologies emerging from the field could profoundly transform society for the better in the coming decades."

Let's be open minded, but careful... these are powerful technologies.

Interview with Dan Rose, Facebook - about how Publishers can make money on Facebook

Interesting long read about Facebook's plans for publihers. It obviously can't be their primary objective to become the most important traffic source for publishers - and then? It is way smarter to enable publishers to make money on FB. Some of the plans are discussed in this interview, among others mid-rolls in videos and larger ad units in Instant Articles.

A/B Testing on Netflix

Everyone agrees that A/B (or more variants) testing makes sense and helps to automate decisions otherwise done by (naturally biased) humans, but I know embarrassingly few companies who have actually implemented this in theor day-to-day operations. While A/B testing sounds pretty easy, it's often complicated to set up and run it in a meaningful way. Always interesting to see how others do it, this post from Netflix's tech blog offers an in depth view on how they do it.