In the money


Yesterday’s Mix Session ‘Web 2.0 and Beyond: What Is the Business Reality?’ would more accurately have been called ‘how are we going to make money out of this stuff?: no new information here’.

Panelists were Bryan Biniak of Jacked, Tim Kendall of Facebook, the ever charming Loic Le Meur, now of Seesmic, and brainiac Chris Saad of the data portability project. The host was Frank Arrigo of Microsoft (his write-up here).

All very promising, but unfortunately it was all over almost before it began as TechCrunch (see their article) asked the first question: ‘Are any of you planning on getting a business model anytime soon?’. A simple enough question you might think, but one that was met with… well… nothing. Loic Le Meur opined that not wanting to know where the revenues were coming from was the hall mark of an enlightened investor these days – palpably a ridiculous comment anywhere but in our current 1999-deja-vu-fest.

Tim Kendall suggested brands should get the audience first and think about money later. Well OK, but how much later? Given that Facebook now definitely has the audience. This is, bear in mind, coming from someone who’s job is to look after monetisation of Facebook.

Loic did add that the options are advertising, ‘pro-‘ versions, and creating a resalable platform. Of course the number one ‘monetisation’ strategy of .coms always has been and always will be to sell themselves to someone else.

Biniak was unique amongst those on stage for having a business that you could sort of see where the money might come from – creating a sort of TV ‘plus’ space for advertising, although that doesn’t take account of how much might have to be paid out to legitimise the content in the first place.

The main reason most people were at the event – I’m sure – was to hear if Facebook does indeed have a secret plan to make money, especially after the bizarrely revealed company’s earnings. Kendall warmed a little in the middle with a couple of interesting ideas, saying the site’s goal was for the ads not to feel like ads because they were so tailored. especially where friends preferences could be re-cycled to encourage word of mouth. There is, he tells us, great click through rates on their ‘social ads’, although he then suggested these rates were ‘almost double’ normal ads – so therefore twice virtually nothing.

I find a great deal of fault with this formula. Social ads might sell me an iPod or the latest marketing text book, but it will not sell me heamorraoid cream. And if I’m buying a car, I will ask my friends myself for recommendations – so how will Facebook get to charge for that? selling is simply NOT all about targetting, and – this is a well rehearsed argument – timeliness is the most important factor in relevance, and when I’m on Facebook, I don’t want to buy cornflakes or whatever someone is trying to sell me (this is why, of course, ad words are so valuable).

After a bit of prodding Kendall started talking turkey. While saying it wasn’t what was happening on Facebook (and a somewhat bizarre  dig at Microsoft which sells display advertising on the site), he suggested that 10c might not be an unreasonable CPM rate for what display was currently selling at on the site. Search by contrast is more like $60. If Facebook could get it up to $1 he argues, that would be a very big business.

The sheer vagueness of this claim seems to me to be quite outstanding. Even if that is is the end game, the sums don’t appear to add up, and in any case, and it is essentially a goal of increasing selling price 10 fold without any visible corresponding strategy.

Chris ended the session on an interesting note talking about ‘VRM’ – vendor relationship management, which is essentially the idea that consumers could get paid to get advertised to. Clearly the detail of getting that to work is incredibly difficult but it remains a lot more plausible than some of tonight’s more whimsical musings.

Monkeying around


If yesterday’s Mix keynote was all about products and developers, today’s was all about Ballmer himself, who was interviewed on stage by Guy Kawasaki.

Kawasaki, an ex-apple evangalist,  pulled no punches – poking Ballmer on Vista, Google,  Yahoo and Apple, as well as some pretty suprising jokes about chair throwing, anti-trust hearings and Ballmer’s infamous monkey boy dance.

All of these were met with surprising good grace (although his less calm side never seemed too far from the surface). At one point, Ballmer even performed a brief  minor variation of his dance on the request of one of the audience members, changing it to ‘I love web developers’. And, at one point we saw Ballmer goofing around and pretending he couldn’t carry Kawasaki’s MacBook Air because it was too heavy, and offering to get him ‘a proper machine’.

The strategy position in general was pretty clear. On the subject of Yahoo, the reason for the purchase it that Microsoft sees search as the killer app of online advertising, and it sees (as we heard yesterday) online advertising as its key monetisation strategy. The greatest ‘synergy’ is pure scale. Does winning mean beating Google? Is it a zero sum game? Yes.

However, whilst Google was the enemy in Microsoft’s online ambitions, Ballmer made it clear that there were other competitors in the other major markets they operate in: desktop, server and enterprise, entertainment and devices. The competitors ranging from IBM to Linux.

One of the overriding messages of Mix08 however, has been that the other fights aren’t necessarily zero sum games, with Microsoft showing a genuine drive to interoperability in many spaces. Are they serious about this? Well AOL and DoubleClick were on stage at the key note, showing how their technologies fit with various Microsoft platforms. It all seems pretty genuine and various people have commented in a palpable change in the way Microsoft is now dealing with the outside world.

Interesting to hear Ballmer describe Microsoft’s Search offering as the ‘little engine that could’, and to point out that the giant was very much the underdog both here and in the personal devices market.

What about Facebook? He made it pretty clear the 2% stake in Facebook was a relatively small deal for Microsoft and that he cared more about their advertising ‘partnership’ with the social network than their stake in it; validating the view that the shareholding is a purely defensive manoeuvre.

The question and answer session came to a close with a very bizarre and slightly uncomfortable question from an employee at Avenue A Razorfish, part of aQuantive which Microsoft bought last year to get their hands on Atlas.

As a lot of people have noted, Microsoft is in a difficult position with Avenue A because of their relationship with other agencies. It’s been assumed it will be run independently or sold off. But the question from one of its staff in such a public forum: ‘What are your plans for us, we hear a lot of rumours’, must still have come as a bit of a surprise.

Ballmer seemed unphased, and said the business would be left to run independently so long as it remained profitable but I wonder if any chairs went flying after he exited the stage.

(More here)

Heads in the clouds


I’m at the Microsoft Mix08 show in Las Vegas with Conchango this week.

There were some interesting thoughts in the keynote presentations that kick off the conference. Apart from some pretty unveiled comments about Yahoo, Chief Software Architect, Ray Ozzie very explicitly talked about the challenges which he sees facing Microsoft in the next few years.

They basically come in just two categories:

The role of advertising

An expectation that advertising revenues online will grow from $40bn to $80bn in the next three years, Microsoft clearly wants to be part of that. That means both creating advertising models of their own and creating tools that enable developers to use Microsoft technologies to monetise their output.

The role of the web

A fairly open nod to the threat/opportunity that ‘cloud’ thinking presents for Microsoft’s traditional software, and increasingly, devices and content marketplace.

He walked through the practical implications of this for Microsoft’s key business areas

1. For personal

The development of connect frameworks (over the net) for keeping devices (‘device mesh’) synched and updated (side loading), and from there, enabling social connections / self expression, from the basis of either entertainment or gaming (‘social mesh’).

2. For business

The drive is from the basic data centre approach, through ‘utility computing’ using virtualisation to distribute and enable services away from the single ‘application’ server approach, and finally to move on to cloud computing.

3. For Developers and designers

Of course the paradigms above – cloud computing from a consumer point of view, cloud computing from an enterprise deployment point of view and developing for a range of different, interlinked devices, will have significant impacts on how developers will have work, how they will learn to design architectures and software.

Ozzie then went through Microsoft’s offering’s across these five areas, although at this point, Vista decided to do a number on me so you’ll have to watch the video for that I’m afraid.

  1. Connected business
  2. Connected entertainment
  3. Connected productivity
  4. Connected devices
  5. Connected development

The remainder of the presentation was led by Scott Guthrie who ran through the features of today’s key releases: IE 8, Silverlight, and some new features and performance improvements to be added to Windows Presentation Foundation (WPF) later in the year.

Dean Hachamovitch (top) came on to present IE. The headlines are that IE8 will make a serious attempt to realign with core interoperability standards, and a step towards HTML 5, as well as incorporating some very interesting atomisation features. Even to the extent that it will render some pages which work well in IE7, IE8 will behave like Firefox and Safari(although a sort of IE7 quirks mode can be forced with a meta tag). This, unsurprisingly was very popular with today’s primarily developer audience who currently have to create various hacks or multiple style sheets to achieve consistent output. Promises of performance improvements (in page rendering etc) also went down well.

The nods to HTML 5 are better handling of the back button for Ajax style interfaces, connection awareness in the DOM (so different action can be taken for offline pages), and DOM storage (the ability to store data locally). All looking good. Also good, although not ground breaking were built-in developer tools.

The next two items however (about 45 minutes into the key note), were more interesting. They’re interesting in part because they provide rich frameworks for ‘atomising’ content and function; and in part because they are very serious efforts by the software giant to create new but open standards (and have been licensed as such).

The first is Activities – the ability for developers to create custom in page functions for acting on in-page content, for example, by looking up a location on website, finding a product on Ebay, searching for terms in Google.

The second piece is ‘Webslices’, a simple mark-up framework for developers to syndicate page areas which then appears in small pop up windows in IE. Nice stuff, and it will be fascinating to see if Mozilla takes it on board or tries to make their own. 

For Silverlight we got a raft of improvements:

  • Adaptive streaming
  • Flexible settings for progressive download
  • Integration with CDN and Server 2008
  • New, skinnable controls
  • Fully working CLR
  • Databinding
  • Unit testing framework
  • Cross-domain capabilities
  • Strong networking features
  • Integration of Seadragon (a powerful image zoom and progressive loading technology – see

We also saw a number of sites who’ve already adopted the technology, most notably a very compelling demo of what NBC is planning for the Olympic Games coverage this year.

The Olympics section is (about 1h20m in) well worth a watch. The new site is set to include picture in picture, live and on-demand across 22,000 hours and 25 sports, visual search and browse, pre-roll ads.

A number of the other features and applications are demonstrated too, mostly in a pretty convincing way.

The last piece was Silverlight on mobile. The announcement being that Nokia will include it on some symbian phones – and it is already part of the mobile platform.

As it always seems to be it’s early days for mobile, but it’s true to say that they have made it work, although details were a little scarce, as were any features of the demo that really added value which couldn’t have been done in HTML.