Canary in Search Coal Mine?

FTD (flower company) preannounced poor Q4 revenues today and blamed increased search marketing costs for the shortfall.  Although it would be silly to extrapolate the experience of a single retailer in a single vertical to the whole of ecommerce, such a data point could potentially signal impending trouble for Google, Yahoo!, and the rest of the search industry.

From the FTD release:

During the Christmas season, certain online search engine costs increased significantly over the prior year, and as such we made the decision not to pursue the resulting high cost order volume. As a result...we have begun making additional investments in our marketing staff to help build a more diversified marketing portfolio.

In other words, in the floral segment at least, keyword prices have increased to the point where one leading retailer has decided that they are no longer economical to buy--and will be moving some marketing dollars offline as a result.  In the short term, this might be good news for Google, Yahoo!, et al (another big quarter).  In the long-term, however, it's potentially a big negative. 

One of the major drivers of search revenue growth over the past few years has been steady increases in keyword pricing--increases justified by the demonstrable fact that ROIs were better in search than for other other marketing spending.  Evidence that keyword prices in at least one vertical have now hit a ceiling suggests that the life-cycle of this growth driver, at least, might be nearing an end.  For the search giants, such a event would be akin to a 747 losing an engine: The plane can still climb, but not as quickly or reliably as before, and passengers must begin to clutch armrests and pray that another engine doesn't give out. 

Comments

[Steve D]

Ebay said too some time ago that keyword prices were out of hand.

It might be that not only prices are not sustainable, but also certain part of traffic volume is under big question - that part of pay-per-click traffic which comes from 1) all those millions of bloggers who display ads on their blogs and themselves click on the ads 2) millions of overexcited individual investors who click on ads just to contribute to GOOG's bottom line 3) those who park domains with companies like sedo.com and themselves click on the ads 4) and who knows what other kinds of abuses are proliferating.

It's good that advertisers can control how much they bid for keywords to adjust to junk traffic to keep ROI at a certain level, but Google will have to deal with that kind of fraud eventually and I doubt they will be able to do it. And the more fraud there is, the lower prices will be, because advertisers will adjust prices to compensate for growing junk traffic

[Henry Blodget]

I agree wholeheartedly. The click-fraud debate will end with ROIs are no longer attractive. Whether this is the result of fraud or just free-market price increases is almost irrelevant: companies will pull back on search spending until ROIs are favorable again.

If, as you suggest, click-fraud is rampant, this pull-back will likely be severe, as it will be accompanied by a general loss of faith in search marketing. At least temporarily, the baby will be thrown out with the bathwater...

[Stuart MacDonald]

This was absolutely destined to happen -- I saw it coming 2 years ago. Why? Click fraud might play a bit of a role, but the biggee is that irrational bidding drives the ROI to nada and big players pull out. No responsible, profitable player is going to bid more than 100% of gross profit on a keyword -- not for very long, anyway. But, because there are yokels who are not bound by any such limits in the short term as they try to drive traffic (and when they go away they are rapidly replaced by a new crop with fresh money or, conceivably, different economics that make higer bids make sense for them) delivered GP for the seller has nowhere to go but down. As you say, Henry, this does not bode well for search players. They had better hope that the fresh crop of irrationals keeps on coming, figure out how to make their natural customers happier with their investment, and figure out how to move up the advertising foodchain, out of straight "sales" and into "brand building" -- the latter, of course, being a big part of the logic behind the GOOGaol deal.

- Stuart

[Steve D]

Brand building is an attempt to blanket out millions of people with ads (often intrusive) hoping it registers with a few of them - highly inefficient method and very hard to track ROI. Practiced historically for a long time - because there have not been technology available until very recently and the very idea of targeted pay-per-click developed even less recently. Now when technology allows implementing all kinds of 'crazy' ideas, advertising is surely going to move away from brand building (although brand building is likely to never die off).

I think what search players will end up with (either deliberately or out of no choice) is having to move to pay-per-sale model. Fraud will be completely eliminated in that case. Advertisers will bid percentage of sale rather than click price. It's much more difficult for search players to implement it (it requires online payment mechanism like PayPal), but it's conceivable. This kind of advertising would compete with Ebay more directly and very likely show that Ebay's maximum 5.25% fee of sale is low (despite certain revolt among sellers when Ebay hikes prices from time to time).

Although, I seriously doubt that current leaders in search will choose pay-per-sale model out of free will. Maybe Microsoft, out of desperation, will partner with PayPal and try it out. If they do, they will make advertisers fall in love with them – no doubt pay-per-sale model with no upfront fees of any kind is much superior to pay-per-click model from retail perspective, and others will likely have no choice but to follow.

[Steve D]

In my comments above, I said I seriously doubt that current leaders in search will choose pay-per-sale model out of free will. But on a second thought, it seems that if retailers allocate certain percentage of their sales to ads – let’s say 10% - regardless if pay-per-click or pay-per-sale model, than total amount of money flowing to search engines will not change because in case of pay-per-click model, as soon as total amount spent, starts to exceed the above 10%, retailer will simply cut off spending – search engines will never make more money than retailers’ justified ROI.

But pay-per-sale model has huge advantage of being 100% predictable for retailers – they don’t need to worry if they get many ‘empty’ clicks and end up with their advertising budget exhausted, but with poor sales. This predictability will definitely make advertisers choose pay-per-sale model. Search engines don’t seem to lose anything in terms of potential revenues, but retailers gain the above mentioned predictability. Given that it’s very hard to establish online payment mechanism, This makes me think that PayPal will be the key candidate to partner with search engines and might even lead to Google buying Ebay (but that’s another story).

And it seems very likely that desperation will cause Microsoft to choose this innovative model, which will be their path to salvation because advertisers will love pay-per-sale. They might charge 1) 5-25 cents fixed per click to prevent 'junk' ads so that advertisers have something to loose in case they put 'junk' ads and 2) percentage of sale, which advertisers will bid through auction mechanism.

[Stuart MacDonald]

Interesting take on what you think brand building is, Steve D. As is your view on a brand's value and the process that you say allows it to happen. Interesting and spectacularly wrong.

- Stuart

[Stuart MacDonald]

Interesting take on what you think brand building is, Steve D. As is your view on a brand's value and the process that you say allows it to happen. Interesting and spectacularly wrong.

- Stuart

[Niki Scevak]

I'd have to strongly disagree here - announcements like these will become commonplace and keyword prices will continue to rise.

It is simple Darwinism. The same keyword is worth a different amount to each advertiser.

FTD may only be able to pay $1 for a certain keyword but the same keyword may be worth $2 to 1-800-flowers (by virtue of a higher conversion rate).

Those who cannot optimize and train landing pages to become more efficient will be left unable to bid on keywords.

Also remember that ROI is a binary concept with infinite demand. Once it is negative it is shut off but up until the break even point people will continue to bid.

[Henry Blodget]

There's no question that some advertisers are more efficient than others and can therefore pay more for keywords. There's also no question that, for many companies, landing-page efficiency, etc., has a long way to go. Based on FTD's press release, though, it seems that search marketing had once been a good channel for them and that a price hurdle has now been crossed that has driven them out of the market. Whether this is the beginning of the end for keyword price increases--or just the end of the beginning--is impossible to say. I do think, though, that the anecdotal evidence that DOES accompany the beginning of the end will look just like this.

[Ian]

Click fraud is a big problem for Google - their going to have to do something to support that massive share price.

[Mahlon]

Looks like FTD gambled and lost. Because they didn't buy search ads, FTD's consumer business tanked over the holidays. The head of marketing got the ax as a result.

http://blog.searchenginewatch.com/blog/060127-000657