Sunday, May 17, 2009
The recent Hitwise report as related by analyst Heather Hopkins should have been vital reading to anyone into digital marketing or search technology. Paid search clicks, the cash cow that almost entirely drives the industry leader, Google, were being portrayed by those reporting on the item as "down 26%" year over year in Q2. Easy conclusion: "the meltdown" has "finally hit paid search."
Economic weakness affects the click auction, no question about it.
But given the potential economic bombshell this represents, it's eye-opening to see the lack of actual responses to these reports. Look at any of the main stories reporting on the item, and they range from the misleading, to the credulous, to the snarky. Cade Metz's snarky response was at least an attempt at analysis. Hundreds of others linked to the story or retweeted it, without it apparently entering their heads for a second that flipping a link up of an oft-retweeted story isn't giving us any more insight than we had pre-re-tweet.
This has actually been a typical pattern. A few weeks after Google's financial reports for a quarter give us real insight into how the business has been going, everyone seems to go back to not understanding how the business works. Speculations on the next quarterly report tend to predict imminent doom, for whatever reason. We're in that "dark, suspicious" period now. We'll wait for next quarter to see if all the weird speculation was based on reality.
Since last year's misleading comScore report that tanked Google's stock before producing a clarification from comScore, it seems that no one's prepared to offer much of an interpretation of third party data at all. So into the void of no analysis, rushes headlines that seem to say paid search clicks are down 26% year over year. Or more accurately, the proportion of overall traffic referrals to (the subset of) websites (in Hitwise's sample), among all sources, that comes from paid search, has dropped by that much.
What can we make of this, really? Here's a stab at it.
1. The data may not be accurate.
1a. Or at least, the data might not be insightful because it treats a click as a click as a click. Some clicks are expensive or meaningful; others are not. There's a whole host of potential reasons for skew.
2. Why are we paying so much attention to travel sites? They do spend a lot. They may be spending a bit less. Their sector has been hard hit. It's hard to say, unfortunately, how the stats on clicks affect the dollars being spent on those clicks (see below). We do know from a study put forward in a webinar offered by Google that travel searches are holding pretty steady, but conversion rates are down. You can certainly see how in a sector like this, with massive overcapacity developing and brand-building efforts in the distant past, dozens of travel sites like travelocity, Expedia, Hotwire, and others, may be scaling back campaigns, and bids. They're one of the top paid click generators, ergo that looks pretty bad for Google on the surface, until you look at the financials overall. Does this mean, then, that the "sky is falling" on the "Google money machine"? That the sky is falling for the travel websites? I doubt it, for either. The much bigger problem is at the end of the suppliers - the hotels and airlines - who are being forced to go to thin and negative margins to fill rooms and seats.
3. The theme that came to light with misinterpretations of last year's comScore numbers has yet to be digested fully by observers: 0verall, the nature of Quality Score keeps more white space on the page more often. (Google's overall monetization rate, % of pages with ads on them, remains relatively low.) Many conventional advertisers are barely affected, though they've been annoyed by arbitrarily high "reserve" prices on less competitive phrases and keywords that supposedly aren't relevant. Many conventional advertisers are still seeing growth in paid clicks; some in weak economic sectors are seeing appropriate declines. Google's purge of affiliate marketers and click arbitrageurs has been widespread, so it removes a lot of low-quality, low-priced ads from the system entirely. Financially this is only a short-term hit for Google. It increases satisfaction with search results and maintains pricing on keywords with high commercial intent. Granted, pricing power is diminished as smart advertisers take this opportunity to lower their bids, but the decline in sheer volumes of paid clicks isn't purely due to a weak economy, but rather, the side effect of a deliberate Google policy that was still going strong throughout 2008.
4. Has Google experimentation with various blended search results and page layouts actually resulted in more people clicking on SERP's, local results, and video content as opposed to looking at paid clicks? This is a pretty bold move on the surface, but owning local business relationships and video ads in the future is part of the overall strategy, so it's perhaps expected that Google will be willing to see users click on the paid ads slightly less often. If it was a shift that terrified Google, I don't think they'd continue pressing forward as aggressively with the changes.
5. Some - maybe many - large-company CEO's are ganging up on Google, informally, as this "don't buy your brand clicks" meme seems to spread. While that adds up to quite a few clicks, those are inexpensive clicks anyway. Brands' refusal to buy those clicks hurts them, in my opinion, more than it hurts Google. Anecdotally talking about "ranking on those terms in the organic results anyway" is not an accurate way to measure the incremental sales lift from those keyword ads, for a host of reasons, including a 2003 view of what it means to "rank in organic". Some folks will continue to try to pinch pennies in the digital channel in spite of the waste going on in other channels; others will get over it and go back to putting more profit in the till with straightforward direct marketing to searchers, including those parts that focus on brand terms.
6. By quasi-organizing against Google and collectively angling to take revenue out of Google's pocket on brand terms (despite Google playing nice by policing trademark in a variety of ways), it appears the Moose Lodge of CEO's may have incurred Google's revenge.
Google may have grown tired of being handcuffed in monetizing those terms by both the rhetoric of trademark holders and Google's own editorial policies. They've launched the first major initiative in the direction of re-opening that monetization avenue, especially in the U.S., starting in June. (Although it refers specifically to ad text, I believe the announcement may also signal a shift in overall stance towards trademark terms.) It almost feels like the Hitwise report spurred Google's timing of the announcement. Insofar as the Hitwise report did a good job of showing that many brand keywords got taken off the board year over year, Google's response is going to be to put them back on the board. IMHO that means not only less editorial squelching of ad copy, but a relaxing of punitive Quality Scores on brand terms. Maybe Cade Metz's verbiage is apropos here: Google's going to "turn the dials" on something they had dialed way back on (largely to avoid the nuisance of legal squabbles and CEO's carping). "Turning the knobs" back towards a more relaxed setting for brand keywords should mean more paid clicks back on the board, more fun and profit for some qualified and creative advertisers, and more revenue for Google.
To put it another way: if you don't want to pay a few dollars here and there for your brand keywords, not only will we let someone else do it, but we'll let them use your name in their ads, as long as it's legal. So there.
No doubt, advertisers and search marketers today need to shift to a more comprehensive strategy of being visible in searches across various search "forms" as search behavior shifts and new blended search formats emerge. But that said, the fact that Google's stock hasn't dropped much on reports of a supposed 26% decline in paid search clicks indicates that investors, at least, now understand that the paid search auction is highly predictable, still a real cash cow for both advertisers and Google alike, and user behavior in clicking on ads remains surprisingly resilient, even in the worst period of negative economic growth seen in most of our lifetimes.
Labels: goog, hitwise
Friday, April 13, 2007
The latest Hitwise figures show Google increasing its monthly share of searches again in March, at the expense of Yahoo (down slightly), and Microsoft (down slightly).
Microsoft's numbers can be attributed to a few glitches, depending on who you talk to. A rebrand of the search offering (bad, confusing idea, I think) or difficulties with Hitwise's methodology.
Netratings estimates a slightly lower number for Google, 55.8% for February. Unlike Hitwise, they seem to assign 5% share to AOL Search. Whichever ratings agency you trust, it's clear (again) that Yahoo is in real trouble of losing its status as any kind of default search box for anyone. That would spell big trouble for the organization as a whole.
So I'd like to focus a bit further on the danger Yahoo faces if they let these numbers slip any further.
* First, they've spent too much on search to abandon it.
* Related to that, they've invested too much in Panama, which was built *primarily* to monetize Yahoo Search and only secondarily as a platform to bid on content, to lose any more search share.
* Third, search is good.
The "typing stuff into a box" thing is too important a category to cede, no matter how navigation may shift in the future. It is not good enough to say that Yahoo has a great diversified model that will make them money from all kinds of ad formats and fees. True, but it's not powerful enough to compete "unfairly" as a real heavyweight, if users stop using their properties to search.
So how to acquire enough of those searches? Big ideas grown internally aren't necessarily the way to win people over from their Google habit. So how to acquire enough momentum to re-establish it as people's habit to search on a Yahoo property for at least some of their needs?
* One way to get some of this back would be to acquire local search properties - like the very hot Yelp. You don't have to get into unfathomable social search or expensive Facebook acquisitions, under that scenario. There are some growing, fairly conventional, properties with a slight cachet of cool that are doing quite well. Get them now, before the price doubles.
* Next, keep building out those verticals. If your leading properties are losing to upstarts, acquire the upstarts purely for traffic. Then get the AdSense ads off them.
* As stated above. Acquire specially selected content sites purely because they're AdSense sites. Revamp the monetization plans of those sites.
* Make sure to internationalize your search for these acquirable properties, of course. Yahoo has certain major international holdings, but they need more.
* Biggest of all: Ask is clinging to some market share, and would immediately add several "type it in a box" type properties to Yahoo's stable. Moreover, IAC owns other vertical properties that would work to reinforce other things Yahoo is doing. Although it could be painful, there's nothing that says Yahoo couldn't launch a bid for all of IAC, sell what doesn't fit, and keep what does. Yahoo's valuation is currently about 4X IAC's. People worry about Yahoo's executive bloat, but the best way to reduce the bloat ratio is in fact to grow your overall top line and overall traffic, as long as it's strategic and as long as a lot of it is scalable search type stuff.
* Try to absorb a handful of departing, cashed-out Googlers who are somehow going to be convinced that this is a really cool challenge.
* Least likely, but a good idea for both companies: convince Microsoft to give up on search and paid search platform building. Re-partner on both fronts.
* Piss off Wall Street somewhere around Q1 of 2008, by implementing a short-term de-monetization plan across all properties to increase user satisfaction and traffic growth. Basically, spend the rest of this year studying how you can monetize your traffic *less*. In verticals, in search, in apps, etc. I know that's already happened in a lot of places, but try to hive off a little more - your effective CPM probably bounces back anyway if you're patient. (See Godin, Seth. The Dip.) Don't think Google didn't just spend the last couple of years doing that after beginning life obsessed with it. The paradox of Google's gentle de-monetization initiatives is that it made them money hand over fist in the long run.
* Finally -- though it didn't work so well for Lycos, consider leaving acquired brands intact. (It works for IAC.) If you acquire Yelp, don't fold it into your overall plan but rather let it maintain its identity for the most part. Use your leverage to distribute it more widely and improve the product.
This post brought to you by the philosophy that launched this site in 1999: while the search & traffic ownership game is not quite winner-take-all, it is something close to that. Monopolistic-type advantages will make it more likely for people to default to your offering. When you own the traffic 100%, profit margin on monetization is superb. Profit margin on brokering media is highly squeezable. There are some searches out there that neither Yahoo nor Google own. It would make a lot of sense for Yahoo to swallow the price tag now, and begin owning them.
Go big or go home!
Labels: ask.com, google, hitwise, metrics, netratings, yahoo
View Posts by Category