More than a tweet, less than a blog.

1982:

via Agile Advertising - Burt at Cannes Lions

Happy midsummer from the west coast of Sweden!

gustafalstromer:

hypem:

(via nathanieljames)

Sometimes the copy and layout of Fantastic Man is just too much perfection.

Of Course Apple Is Going to Do Search. - John Battelle’s Searchblog

Rafer sez:
Nope. He also wasn’t interesting in tablets, remember? It’s just yet another remark that means, “don’t you know I deny exactly what we are building now?” It’s HTML5-specific web search — and he’s already got the ads for it.

(via rafer)

Teddybears vs Don Carlos at Dramaten. Very, very awesome.

An interesting discussion has emerged regarding how the ideas of bestselling baseball book “Moneyball” can be applied to advertising. ITIBTWIC started it, Ad Contrarian continued and Dave Trott locked it down.

Moneyball has been sitting on my bookshelf since forever and for whatever reason I have yet to open it, despite the fact I really like other stuff Michael Lewis has written.

Anyhow. The idea behind Moneyball is that statistical analysis can identify players that drive success, which is not necessarily the same thing as being famous. If you can poach these key players that others haven’t identified as critical for a discount, you can build a winning team for less. It’s data driven cost-efficiency at it’s finest.

Applying this idea to advertising is an intriguing thought, since agencies share many similarities with sports teams - people are it’s most important assets, with different roles required to reach the intended result, they’re very competitive (internally and externally), with little institutionalized knowledge etc.

Dave Trott makes a good point that in order to make a good statistical analysis one must agree on what variable you’re measuring. Baseball has one simple variable to keep track of - the score. Advertising on the other hand has many different ways of measuring success - awards, brand awareness, intent to buy, positioning, ROI etc. 

However, I think that what Dave is missing is that if you run an agency, that variable is simple. Sales. The logic is simple, if you can identify the individuals driving sales at agencies and poach them for a discount you’ll earn more money. Then again, your objective might not be to build a huge business, but don’t worry. Gunn Report points works as well, it’s all a matter of you deciding on a single variable to build your algorithm around.

Though I usually agree with Dave, I don’t think that it’s a problem om finding a variable that reflects your priorities. It’s a matter of finding good data sets - where on earth would you find a well structured, high integrity data set for who worked for what agency, with what team, for how much, coupled with the agency’s P&L statement and award record for that period, I’d be more than happy to spend a weekend on data mining and see what I find.

Bring it on.

(via loveyourchaos)

Stamen’s latest. Awesomeness.