Customer Rotation

Paul Kedrosky takes a quick look at a graph of advertising growth and GDP growth since 1983 and says, essentially, "See, they're correlated." Well, yeah. But he's not answering his own question: how does this affect Google/Yahoo?

Ad spending growth is highly correlated to GDP growth. About 52% correlation since the end of World War II, according to my data. And, from a quick peek at my graph, it looks like the other 48% is often an exaggeration of the GDP movement.

Note, though, that there have only been three times since WWII that ad spending has not grown (in nominal terms): 1961, 1991 and 2001. Couple this with the fact that even if ad spending is flat, money is moving from offline to online, so online will continue to grow. Kedrosky just isn't making the point he thinks he's making.

But I have some reservations about Google myself. While Google is a good proxy for online advertising in one sense, since it's so large a piece of the online advertising spend, there are some limits. Online ad spending is not all one thing, and one type of online ad spending can come at the expense of another type.

So, here's what's been bugging me: (a) I recently spoke to a good company that sold online mortgage leads to brokers that has had to shut down because their customers stopped buying, and (b) Niki Scevak points out that LowerMyBills just cut their affiliate payment from $40 to $6. This suggests that mortgage lead gen is hurting.

Okay, no need to snicker at me. Here's the puzzling part: take a look at the Mortgage Banker's Association's data on mortgage originations applications.


Look at the graph for last year and the beginning of this year: mortgage applications are doing just fine, while mortgage lead generators are hitting the skids. How to explain this?

Here's my hypothesis: advertisers are fleeing to a different type of "quality." They're moving to display ads, they're trying to figure out social media marketing, they're building widgets.

Normally in a downturn, there's an increase in spend in measured media, the more measurable the better. Downturns feed direct response. But this downturn, if it is one, is different. It's not driven by fundamental weakness in the economy, it's driven by a rotation of customers. Mortgage lenders still have plenty of customers to choose from, but they've completely reversed course on who they want as customers. A year ago, subprime borrowers were the most profitable; now a lender wouldn't touch one with a ten foot pole. When you're trying to bring in subprime you use Google, you use email, you use targetted, transactional, response-driven ads. Lead generators are the kings of this kind of marketing. But when you want more of the everyday, safe but low-margin customers, you don't need targetting. You need reach and frequency and all that. This may be bad news for Google (relative to its past growth) but is probably good news for some of the more mundane but more established online media.

Another Opinion on Financial Services Advertising

AdAge has a roundup of what various marketing professionals think will happen to advertising spend in 2008. Of the ten people they quote, only one is a relatively impartial observer: TNS' Jon Swallen. What does he say?

"The lesson to be learned from the past few recessions is that it obviously impacts different categories differently. Automotive ... is already pretty grim, and budgets have already been pruned across the board. Retail is traditionally the most economically sensitive category when you have a downturn in GDP. We're already seeing a flattening in spending by department stores, restaurants and a host of categories related to housing (including hardware, home furnishing and appliances). ... What's kind of interesting is that financial services -- which is sort of in the center of the storm -- has actually been quite robust. In the short term, at least, these guys are battling for a share of a shrinking pie."

The last two sentences sound familiar.

But I have to admit that I'm surprised by his take on automotive advertising. His opinion does not seem supported by the data. I tend to side with the--admittedly impartial--CMO of Harley, who says "Our belief is that spending through a market downturn creates competitive advantage for the market upturn."

The New Universal Identifier

Josh commented on my last post that privacy requires control over both (a) personal data and (b) the platform on which the data is used. My gut instinct is that our expectation of privacy is only the former.

Privacy is a moral issue but, as with most moral issues, the ideal compromises with practicality in our reasonable expectations. We can reasonably expect that personal data under our control will be kept private. We can reasonably expect that if we give personal data to someone else and they promise to keep it private, that it will be kept private. But I don't think we can reasonably expect that if we give personal data to someone else and all they promise is that they will make it kind of hard to get to that it will be private.

We've discovered this over and over: with real estate records, campaign donations, Google Street View. The probabilistic idea of public actions being effectively private by being lost in a sea of noise, being alone in the crowd, has been confronted with technological reality (although, as with most erosions of privacy, we have been slow to notice or meager in protest.)

I don't think this is a good thing, I just think that the expectation of this sort of privacy is so unrealistic that it's not worth complaining about. There are ways to use technology to counteract the loss of privacy from technology, and that's what we should be asking for. In this case--where our email addresses are used as universal identifiers, the new social security number--we should complain about any service allowing someone to see our email address at all. Given the way things are going, it won't be many years before you can register for a driver's license with just your email address as ID. Email addresses have become our new true names, and we shouldn't let anyone but our trusted friends have them.

Drawing the Privacy Line

Greg is bothered by Plaxo scraping your friends' email addresses from Facebook. But I think he's less bothered about the automated downloading of them from Gmail and other services. Paul Buchheit points out that the latter is just as much a violation of Gmail's TOS as the former is of Facebook's.

Why is doing it the hard way more of a violation of privacy than doing it the easy way? Just because Gmail has an API for it and Facebook tries to hoard the data for itself, is there a difference?

The way I see it, if your friends are making their email addresses available to you through Facebook, then trying to put those email addresses into a place where they are easier to use isn't a violation of privacy. It would be a violation of privacy if Plaxo made your friends' email addresses available to anyone else, of course.

Am I missing something?