Social media for the Paranoid

Blow torch and crème brûlée

Wednesday night at the Twitter Chirp conference was a time for much gnashing of teeth as the assembled flock swooped around the San Francisco docks and crapped all over Twitter’s catering: There was no beef. No chicken either for that matter. No even a prawn cocktail. Believe me – I asked a caterer and she looked at me as if I was asking where the fresh blood was on tap (we were a long way from a full moon so she had nothing to be worried about).  It was Vegan, and that was how their new revenue model appeared.

By the following lunchtime when the catering was provided a cluster of lunch vans – featuring food like crème brûlée and spicy beef burger – and the ensuring competition provided what people really wanted, the impact of the revenue model became clearer.

I believe that Twitter’s revenue model will work – and will have the opposite effect from what is feared by their entry into the application space – namely that their share of the display of Tweets will drop from 25% to around 5% as their user base increases by a factor of five. However the good news for them is that they will keep the beef while some of their partners will get the bread, and the others will have a blow torch applied to their crème.

So the basic idea is that of ‘promoted Tweets’ – Tweets that you have to pay for. To understand this you really have to understand that Twitter is a near real-time high velocity broadcast medium. What on earth does that mean? Essentially a small number of topics dominate the Twitter conversations at any one time. As I write one major topic is the grounding of European flights because of the Volcanic dust in the jet stream. Many people monitor all the posts related to a topic such at this by using ‘hash tags” essentially special search words that people add to their Tweets. So for a big event like this there are thousands of Tweets a minute, and anyone post disappears off the screen very quickly. So if you want to tell the world that you will be running a special bus from London LHR to Paris CDG airports for stranded passengers it is very unlikely that many people will see the message. Hopefully some people will see it and “re-Tweet” it to their “followers” (essentially people who elect to follow what they write) and so the message gets out. Promoted Tweets stay at the top of the list of message for longer. How long is not clear but lets not go there for now. Lets just assume that it works and people click on the promoted Tweet and get on the bus.

The bus company is going to pay Twitter for this – what this is (user impressions, clicks, etc.) is not clear and that really does not matter for now.

What matters is that Twitter will pay the company that displays the Promoted Tweets to the end user roughly 45% of what they get (after their direct costs are deducted and the remainder split 50:50). I’m going to call them “Display Companies”.

So at first pass this looks like Google’s syndicated network where Google Ads appear on user’s pages and the users get paid for displaying them. But there are several crucial differences – the first is that Twitter is providing both the Ad and the Content to the Display Company. The second is that Twitter displays a minority of Tweets (25%) and their Display Company partners (everyone from Google to a one man development shop display the rest).

So for the Display Company they don’t need to generate their own content, instead they need to add value to the content being provided by Twitter. I am now going to argue that for pretty much every good use of Tweets it will be better for Twitter to let a third party handle the Display.

Editorial Control

Twitter is essentially like a cellular company that transports text messages (SMS) – they are covered by the Common Carrier laws (or similar laws in other countries) which basically mean they can not be sued to the basis of the message that they are transporting. The same is not true for the Display Company if they do too much processing on the messages. How much is not yet clear, but clearly the more you mess with the flow the more value you add and the less like a Common Carrier you are.

Twitter either does not see this distinction, or they think that there is enough business in simply displaying the Tweets without exercising any editorial control. If it is the former then they will learn fast, and if it is the later then there is a huge opportunity for other companies to act as “front men” – clearly not big visible front men, but thousands of small operators in difficult legal regimes (think Venezuela).

Mashup

Then in the space of unmoderated display the biggest value of Tweets would seem to be in concert with other information. So returning to the example of the Volcanic dust in Europe there is alot of value in a “mashup” that combines a map of the affected area overladed with the intensity of relevant Tweets and a travel information service, such as Kayak.com.

Mobile

At the moment Twitter is making a major push to have their application running on every mobile platform.The stated goal is the make Twitter more accessible – and reduce the hassle of selecting and installing a third party application. Facebook does this, so it seems reasonable for Twitter to do the same. It seems to me that this is an intermediate state of affairs and pretty quickly this functionality is going to be built into the base operating system of the mobile phones – just like the case for Texts (SMS). There are advantages all around – Twitter gets massive penetration and a barrier against competition, users get an instant use experience, the cellular network operators get a slice of the Twitter advertising revenue, and the phone OS providers get another feature in their feature wars.

Search

At the moment 25% of Twitter users go to the Twitter website to search for Tweets. Going to Google or Bing allows them to search Tweets and other content, so it seems hard to believe that they will continue to conduct search on different webpages. For complex Tweet search, such as multiple simultaneous, real-time search they will continue to use Twitter clients such as Tweetdeck.

Conclusion

Twitter has to choose between control and growth. If it is to reach its stated goal of 500 million users (time-frame not stated) then they can’t control the Display of Tweets, and they will need to stick to the meat of the problem  (transport and search) – while their partners focus on the bread and butter business of display. Incidently this conclusion is also true (for different reasons that I will explore another time) for the creation of Tweets.

So give it a year or two and Cisco will buy Twitter – after all they bought Webex.