Cogblog

Apparently, AIG is “too big to fail“.  This made me wonder: What is “too big to fail”?

It makes sense to me that if we are going to declare that a company is too big to fail, then if we are going to introduce new regulations, the first regulation we introduce would be that companies cannot become “too big to fail”.

I think it is safe to assume that when they say “too big”, they don’t mean market capitalization.  If they did, then Google went from not existing to too big to fail in record time.  I doubt they mean profit as these financial firms results can swing wildly year to year (AIG had $18.5 billion in losses the last three quarters, so their profits were clearly not being protected).  If they meant customers then McDonalds or Starbucks would be too big to fail and I think we can all agree that no one bails out Starbucks.  So we must mean revenue, right?

Let’s look at the companies we need to break up if AIG is “too big to fail”.  (Although obviously most of these companies promptly relocate to Switzerland, so that would have to be figured out.)  AIG had $110 billion in revenue, making it a top 20 company on a revenue basis, so the list is actually fairly short – and the fact that this is true maybe implies that if it is possible to be too big to fail than maybe AIG is!  Maybe the legislation would draw a clean line at $100 billion in revenue?

• Wal-mart – hard to figure out what breaking up Wal-mart looks like.
• Exxon Mobil
• Chevron
• General Motors
• Conoco Phillips
• General Electic – easy to break up
• Ford Motor
• Citigroup
• Bank of America – hmmm
• AT&T – breaking up AT&T would be funny, right?
• Berkshire Hathaway – easy to break up
• JP Morgan

After that, HP and IBM both appear on the list with ~$100b in revenue.  Once again, break-ups that would trouble no one.

Frankly, Wal-mart is the only company on this list that I think it would be strange or bother me some if they were broken up.

Always appreciate it when someone takes me up on an offer and the result was that I had a super fun dinner last night with Greg (one of my favorite bloggers), Emmy, and Azeem.

Greg let me rant and rave most of the night about behavioral targeting and I came out of it thinking, “that was only helpful to Greg if he wanted to feel like that was never going to work.”  I spent some time after that, as Emmy would have said, “trying to think about something nice”, and I had a few realizations:

• It is very hard to understand the value of behaviors, but that doesn’t mean they don’t have value.  Personally (i.e. my gut instinct), I would rather show ads to someone that I know visits finance sites every day, regardless of where they are at that moment, than someone that is on a finance site right now.  So a behavior that shows consistent interest + random current inventory > expensive inventory without a guarantee of true interest in a subject.
• Great optimization, the best thing an advertiser could possibly have online, can theoretically happen two ways: Awesome algorithms; or OK algorithms combines with awesome data.  This is well-known.  So advertisers want what behavioral data exchanges are selling.  That has to be good!

Greg can also now testify to the awesomeness of Cogschwag, so enter my contests and win some prizes already.

I know its the political season, so it is all sound bites, but AdRoll has piled up favorable press in a way that I can’t understand.

So they are building ad networks but only selling them through their self-serve interface.

“Adroll assembles blogs into co-ops and promises to pay more than the average $2 CPMs that many sites in the middle get.”

“Unlike Federated Media or Glam, Adroll doesn’t go out and try to sell ads for the blog co-ops that have signed up with it. Federated Media and Glam charge their publishers a pretty penny for doing that (about 40 percent of the revenue they bring in), but they do come back with high-paying ads, typically $10 CPM or higher. By automating its offering, Adroll will serve ads that typically pay publishers less than that, but much more than the $2 CPM typically paid by your average ad network.”

Now, I will say that if your point is just “we will sell what we can and sell it at high prices, ensuring that if we actually sold something, you would make a lot of money”, that’s fine.  How do they drive traffic to the site?  Why will people shop there?  I can’t make heads or tails of that.  If you don’t push inventory, you never make money.

Furthermore, I went and looked at one representative network: Alt Music Community.  Why that one?  It was the first on their list.  Anyway, here are the problems I see:

• The majority of the traffic is international.  That means it is basically worthless.
• Tons of below the fold impressions in small sizes – the one site I looked at was AudioPorn Central – I loved the name! – and the 160×600 was 6 page dn’s down.

Is that worth $2.53?  (or $2.37 if you agree to take all those international impressions)

It’s a hard sell.  Especially with no frequency cap.

I will say that I like how they show you where on the page the ad is and they show you impressions by slot by publisher.  If they showed you ctr by slot – while there are data problems with just that raw number – it would go a long way toward helping advertisers value inventory appropriately.

I have been thinking it would be great if I went back and pulled some of our enthusiastic emails, but then I thought, “Why not make it a blog post! There have been too few of late!” So lo, there was a blog and a post was posted upon it. Prizes were found!

We are looking for some sweet testimonials to help prop up our guerilla marketing activities. In the comments below, give me your best stuff.

Here is what we need:

• A good two sentences praising or despising the product, however you are feeling.
• Your name/Company
• A URL if you want us to link to it

Remember how success stories sound: “Before we had a terrible problem, but then we used Cogmap and now we are happy.” Keep it lively and active, keep it fun, and details make it better. Facts & Figures > *.

The most awesome testimonial will get a “Cogmug” and some stickers in the mail!

Thanks for the support!

Lots of travel coming up.  If anyone wants some super sweet “removeable laptop” stickers, leave a comment or shoot an email to brent at cogmap and we can get some coffee.

• NYC – Sept 10
• SF – Sept 13 – 22
• NYC Sept 23-24 – MIXX/Ad Week
• NYC Oct 7-8
• Baltimore Oct 16 – I am speaking at ignitebaltimore.com.
• NYC Oct 22
• Baltimore – Nov 1 – SocialDevCampEast
• SF – Nov 5-7

Busy month!

ComScore has gone to a lot of trouble to demonstrate that their data is better than Google’s AdPlanner.

What comScore misses is that they are falling into the classic Innovator’s Dilemma here. Yeah, Google’s data is bad now, but what about 6 months from now. What about a year from now. When does it become “good enough” for 10% of comScore’s customers? 50%?

Everyone hates comScore. They overcharge for their data, prevent people from looking at custom slices, and service is just ok. A free competitor that gets close will devastate their business. They claim they are safe because their data is better, but we all know that it is wrong too, just less wrong. If comScore is 90% correct and Google is 75% correct, comScore does not have a sustainable business. That’s a fact. Also, as you approach 100%, it gets harder to gain a point of accuracy – i.e. it is harder to go from 99% correct to 100% correct than it is to go from 50% correct to 51% correct. Google will catch up quickly. What innovations will comScore bring that will increase the value of the product? They will have to be pretty exciting because Google is going to take the low end of this market away incrementally and if comScore is not delivering huge value above and beyond this simple accurate data, they will find themselves without enough customers to have a valid business.

Search my wordpress archives for all my Innovator’s Dilemma posts. I am too busy to find them all and get the links flowing today.

Hipmojo has an interesting article that asks the question is content the new king?

One point he makes is that excessive distribution dilutes the value of good content. This is certainly true, and yet there is very little content so valuable that hoarding it benefits the content owner. Information wants to find distribution and people that don’t syndicate content risk having other people syndicate competing content, slowly devaluing the overall genre of content.

If Yahoo didn’t syndicate its news, someone else would.

Furthermore, I suspect we will find out pretty quickly that there is very, very little content so good that strangling distribution benefits the content generator.

I think the benefit of an ad network, relative to a publisher, is that not owning any content, simply buying temporary access to distribution, prevents the network from being held captive by their own mores. A network moves from content to content, following the demands of the audience. Because they treat content generation as a commodity marketplace, they can move more nimbly to follow audiences. Publishers are in some way restricted by the very content they generate to the audiences that consume that content. If social mores change, their audience changes for better or worse. They must adapt their business to follow that audience. Ad networks never gain the encumbrance of needing to adapt.

Of course, a savvy publisher would argue that an ad network is forced to pay a premium for access to premium audiences at all times. A strong publisher that is in an appropriate market could theoretically generate outsized margins by arbitraging their content and maximizing its value in the marketplace.

More later.

Great release of Cogmap tonight. We now show a list of people that subscribe to maps.

More importantly, you can now upload CSV files to maps. These people are then added to the unmapped section of a map and you can begin moving them around.

This is a feature people have been asking for since we launched, so I am pleased to see it get out the door. If you have any problems, don’t hesitate to email us (please attach the file you are uploading) and we will take a look.

Update: Of course, we broke a bunch of stuff in the process.  I think it is all fixed now.

Amusingly, Matt Marshall goes on a VentureBeat rampage with some slam dunk reporting that I feel compelled to discuss.  I will try to note where I am simply applying common sense and where I actually have a bit of knowledge related to the on-going proceedings.

Paid Content discusses the matter fairly well, including an interview with Lynda, but I will get into that as well.  VentureBeat updated their post to point to PaidContent on 1am on August 4th.  Good to see Matt working late, but sad to see a mangled story.  Everyone cites a ton of un-named sources.  No idea who these people are, but it is pretty funny to see these sources used to justify uneducated statements.  I have to imagine his sources, assuming they are good, cringe at the way their comments are misinterpreted.  A great example of reasons to not go on the record.

Matt is obviously super-confused about the roles of publishers and advertisers in networks.  Lets break it down:

The entity, recently bought by AOL, is instead being folded into AOL’s Ad.com/Platform A division. This is a shocking move for some, because Ad.com doesn’t target much at all, and offers ads of $1 or less per a thousand views — and is generally considered a “bottom-feeder” by some in the industry. The company apparently hasn’t communicated very well about what sort of targeting technology Ad.com will offer to the affected customers (see memo below).

Hello, there is no targeting here.  This is buying inventory from publishers to show ads on.  Ad.com buys more inventory than anyone.  They are the biggest network around.  The result is that they pay less.  Volume discounts happen in this industry.  Generally, Ad.com has a strong performance focus and tends not to pay high CPMs for inventory.  Lower prices improve back-end performance of direct response activity.  When someone buys inventory, the kinds of targeting they offer advertisers isn’t really relevant.

I will say that I have knowledge that Tacoda is being integrated into Platform-A, as are all of AOL’s recent advertising acquisitions.  I will also say that Ad.com targets a lot.

The next paragraph is where he added the update referring to PaidContent and Lynda’s interview:

She says Ad.com was targeting the same way Tacoda was, and thus the implication is that Ad.com was paying as much. But that’s not what we’ve heard from sources. Only time will tell.

The sources are probably right because that is not exactly what she said.  She said that Ad.com offered similar targeting.  Probably, advertisers pay the same kind of CPMs for similar targeting.  This has nothing to do with buying inventory.  Advertising.com buys a lot of inventory from publishers and is able to monetize non-behaviorally targeted impressions fairly efficiently.  The result is that Ad.com can buy every impression for sale, not just ones that contain behavioral data.  This allows Ad.com to pay less for your average piece of inventory than someone that is only taking very specific, very lucrative impressions.

Back to Matt’s original article:

Tacoda became popular because it offered an industry high rate of $2 to $6 per a thousand views (CPM).

Publishers love networks that pay a lot for inventory.  But what was Tacoda’s fill rate?  They bought a fraction of the inventory that Ad.com buys.  That is great if you are the publisher… why not sell 10% of your unsold inventory for $2 instead of $1?  Still, if you were Platform-A, you are simply competing against yourself by continuing to pay these rates for inventory that you probably get anyway.

Matt concludes with:

A letter sent out by Tacoda today suggests publishers will have to sign new contracts, and will have to change tags on their web pages.

He then includes the text of the letter, which says, “take down our ad tags because we aren’t buying inventory, but if you want to remain in our data network, you do not need to change your tags at all.”  Seriously, that is what the letter says.

Not sure why Matt wanted to write a grind article on AOL, but it really seems like he felt like breaking out the hatchet.  Was he being fed by an ex-Tacoda person?  I certainly thought so when I read a quote like this: “Tacoda, all automated, paid rates well above other comparable “remnant” networks. It did so by working hard to target behavior of Web users.”  The idea of the company working hard seems to me unsupported fact, although everyone I worked with at Tacoda worked very hard.  Much like people at Platform-A.

Anyway, the one thing I thought was weird in the PaidContent article was this “source”:

One source presented this view to me earlier today when asked about AOL’s decision: “Given the fact that publishers must take all [Advertising.com] ads to get any behavioral targeting/Tacoda ads—with Ad.com selling all the ads—it sounds like AOL has abandoned a large, fast-growing and very profitable market all in the name of functional alignment with a slower growing legacy business.”

Is he saying that publishers only want really high paying ads?  Certainly true, yet my impression is that inventory, particularly arbitrage-able (read: inexpensive inventory (frequently social media) that contains valuable behavioral data) is growing more rapidly than it is being monetized today.  Advertising.com has great reach now.  They do not need to pay a premium to cherry pick behavioral impressions.  They see everyone already.

If Tacoda and Advertising.com were bidding against each other for inventory, I have no doubt the media would lambast Platform-A for not leveraging economies of scale.  With a single inventory acquisition team, it seems like they are being smart.

I know this kind of thing happens in the industry, but I can still blog about it.

The Online Publishers Association (OPA) revealed a study that showed that members of the OPA have better performing inventory than other people.

True story.

K, that was kind of mean, but it felt a little Onion-esque to me so I had to roll with it. In particular, they single out portals and ad networks for their righteous slaughter. As always, they don’t give us an excel file to download to do our own analysis of the data, so I have a bunch of questions and problems. I won’t drill down into the “interactive & rich media” ad analysis they did, because my points are applicable there as well (in fact, more true!), but you can read this and then read the study and see all the same issues

“Ads on Content Sites provide double the brand favorability and purchase intent than advertising placed with ad networks.”

1. So it seems like Dynamic Logic just aggregated every ad network study they have ever done here. Can we assume that we are basically talking about something that is maybe, MAYBE a slight premiumfrom a price/inventory mix to a RON ad network buy?
2. So if ad networks ran 252 campaigns for the sample, and OPA members ran 1,185 campaigns, could one assume that OPA inventory tends to be viewed as more brand friendly inventory already?
3. MOST IMPORTANTLY, so if we assume that the average ad network buy ended up being some slight premium to RON, maybe $1.75? (Really good networks might charge more for RON, but this is an aggregation of all networks, which means it is a lot of inexpensive UGC & Right Media), then is the inventory the OPA was selling sold for less than $3.50?

I have heard, and this is just hearsay, that apparently it costs more than $3.50 to buy inventory on ESPN, Forbes.com, CNN, and Cnet.

So even if the results of this study are true, if the advertiser has to pay more than $3.50 then the buy is inefficient. The advertiser is over-paying for the performance lift.

Fascinatingly, when PaidContent covered the OPA press release, Jim Spanfeller, the CEO of Forbes.com jumped into the comments and said (and didn’t say):

“…This study shows clearly that the premium environments offer substancail value above broad based non premium plans…”

A few things jump out:

1. As a member of the OPA, his comments are in-line with the assumptions (including the flaws) that the OPA makes. The value delivered by premium environments is not necessarily in line with the cost to acquire the inventory.
2. Despite his OPA membership, he does not talk about what kinds of inventory comprise a “premium environment” and that is probably because Forbes.com is starting their own ad network. There is no doubt in my mind that Forbes positions their network as “a premium environment” despite the fact that very little of their non-Forbes inventory probably comes from OPA publishers.
3. So one thing I took away, implicitly, and he would probably never admit publicly, is that he may secretly believe that a well-targeted ad network campaign can perform just as well as premium inventory. Which kind of makes sense given the outcomes of this study. Could an ad network campaign with social media inventory perform twice as good as these RON studies? For sure.

Funny stuff.

(As always, these are my opinions, not my employers)