At SES San Jose, Nick Fox, Google’s business product management director for AdWords tossed out one potential vision of the future. Aldous Huxley would have been proud. In this vision, 5 or 10 years down the road, advertisers won't have to specify keywords on which they'd like to advertise, nor would they have to pay for traffic that doesn't convert. Advertising would be keyword free and paid for based on performance. Sounds compelling! No keyword construction and maintenance, no worries about tying traffic to the right landing pages, no complicated bid management: Google will handle it all for you. Heck, they might dynamically write the ad copy that most resonates with each individual user! The advertiser just sits back, relaxes, and lets the good times roll. Google would be wise to think this over carefully. Three considerations come to mind:
- Could they really pull this off? For this to make sense, it would have to be more profitable for Google to pull all the strings than to allow advertisers to have the controls. For that to be the case Google would have to be able to do the advertising more efficiently than the advertisers do it for themselves, otherwise advertisers would not be willing to pay more for the same results. There are many reasons to question whether that's plausible.
- Would Google be quicker to recognize stock outages than the retailer is? For companies selling shoes, oftentimes the issue isn't an outage as much as thinning of the most common sizes. For airlines, will Google know when certain routes are close to capacity for the likely selling window?
- Would Google know a retailer's shipping cut-off for Christmas delivery?
- Would Google know when a retailer is having a promotion on certain products or across the board? Advertisers can anticipate these events and bid proactively...Google couldn't.
- Would Google know which products generated a great number of returns?
- Not all orders are created equal: leads vary in quality, some keywords generate much bigger ticket orders than others, and the retailer has different margin structures built into its sale prices. Would Google allow different CPAs based on these factors? How would that work? Would Google have to know everyone's profit margins on each product? Would they wait for leads to close to prove the value?
- Would Google be able to pick the right landing page for each search? They don't do this well in organic search at this point. Do a search for "nikon digital lenses". What I see is a number of landing pages for Nikon products (not specifically lenses), and a handful of specific product pages (too deep). This hurts conversion rates. If Google generates lower conversion rates because they pick the wrong pages (could they even find search results pages?) how could they be more efficient than the advertiser?
- Would the advertisers want this, even if Google could do it profitably?
- For this to work, Google advertisers would all have to give Google real time conversion data. Privacy concerns aside, anyone familiar with selling through Amazon knows the legitimate business risks of sharing sales data with a mega company that could turn the tables on them.
- Given the likelihood that Google will have to do a great deal of testing to determine what traffic to send where, and that experimentation by definition involves sending folks to the wrong place sometimes, are advertiser's likely to be okay with the impact on their brand?
- Anyone who works with affiliates will recognize the perils of revenue-sharing deals: does Google get paid for brand sales? How would orders that touch multiple channels be handled? What about frauds, cancels and returns?
- The problems with revenue sharing deals go deeper than that. They create incentives to grab only the lowest hanging fruit from each orchard before moving on to the next. In this context the fruit left behind might represent the long tail: note that when Google Campaign budgets are applied, an already complex puzzle gets even more complex for Google. They simplify the problem by ignoring very specific keywords and just serve the general stuff. This hurts conversion rates. Or the fruit that gets ignored might be niche advertisers. "Too complicated to figure out which of 500 retailer's ads work best for "Sony 60 inch TVs" -- just serve ads for Best Buy, Wal-Mart, Amazon, etc and forget the little guys."
- Google would live in the world of revenue per impression for each ad/user search. If Google sees the same revenue per impression from multiple advertisers, they're ambivalent as to which ads get served the highest, or make it to page one. The advertisers are decidedly not ambivalent. If Google decides Apex ads generate the same revenue as Acme ads they get to pick who wins.
- Would Google want this to happen?
- Given that Google might still control the lion's share of the market 5 or 10 years from now, do you think the SEC and the DOJ might be interested in how Google decides which ads to serve? If the advertisers don't control their own fate, it seems like Google gets to play king-maker, and that seems likely to incur the wrath of federal regulators.
- Does Google really want to move away from being a technology company towards being a service provider? If they think they have client service issues now, wait 'til they're controlling ad service! Yowza will they need to expand staffing!
Join the Discussion