Google’s Little Game of Show and Tell
|
Written By Reprise Media | September 28, 2005 | Share This
|
|
It’s no great revelation that the search engines and their different technologies have become increasingly nuanced, complicated and inadequately explained. What we have with the search engines has become a game of diversion. They’re snapping their fingers up high, drawing attention to new product releases while distracting advertisers from what’s happening down below, developments that speak to the constant evolutions of older flagship products.
One example, Google recently switched over from their Smart Keyword Evaluator system to the Quality Score indicator. They did a great job informing advertisers of the change-over and its implications, complete with online demos and in-person sales calls. What’s been conveyed in the product launch is that the Quality Score no longer draws a graduated system of keywords moving from ‘on hold’ to ‘in trial’ eventually to stable delivery. The Quality Score also provides an opportunity to pay CPCs below the prior minimums of $0.05, and the system will now tell you what bid would be required to reactivate a keyword that’s been disabled. That is extremely helpful. But we’re also told that Quality Score can be applied at the keyword, ad-group or campaign level…problem being, we don’t know how the three work together.
More recently we’ve encountered some problems with a particular keyword that should largely fly beneath the radar. Reprise Media recently launched a proprietary product named “FeedCast” and attempted to buy keywords related to this product, most of which included the trademarked name. Each day we re-launched and by nightfall we were pulled. We kept raising the CPC to whatever Google suggested (which eventually reached $0.50). Our Google rep told us that since nobody else was buying the keywords, even if we conceded that full $0.50 max bid, we’d still pay no more than $0.05 per click should a user actually click on our ad. Well, it turns out that we did get clicks, and the system did charge us the full $0.50 max bid.
So what’s going on here, is this one of those sketchy business practices people always accuse Google of, a collusion of process and technology to inflate their bid rates even further (for another example, play around with the max bids suggested by their inventory forecast system…)? Or is it simply an imperfect product launch with a couple of kinks? We know that Google won’t necessarily serve an ad for every keyword, in cases where none of the advertisers meet their Quality Score demands. But here’s a keyword that’s seen fewer than 1,000 impressions. As usual, there seem to be algorithms at work here, and the nature of those algorithms is alluded to, but never fully explained.
Another example might be the unpublicized expansion of Google’s broad match to now include certain matches more similar to Yahoo with their Advanced match functionality. It’s been known that “tennis shoes” on broad match might include for an interpretation like “ideal shoes for tennis”. Broad match allows the keywords bought to show in the resultant match, but with additional verbiage and not necessarily in that order. The new interpretation will match up “tennis shoes” to “tennis sneakers”. In a less obvious case, we’ve found “fall clothing” to equate to “fall fashions”, which seems a little less direct. “Fall” is the common denominator, and that’s the term that’ll be served in boldface within your ad. The un-boldfaced term of “fashions” may simply confound you, if “fall clothing” is the keyword you were reaching for.
Not to single out just Google, we’ve known for a while that Yahoo has a matching feature operating beyond Matchdriver, this technology looks to monetize as much traffic as possible and might match a search request against a keyword where more than 50% of the terms are included (provided no advertiser bids for something closer to that search request). Some of the more alert SEMs may have noticed that at some point Yahoo delves into the creative or domain to create a match where there wasn’t one before. Is Matchdriver described? Yes, conceptually. Do they ever allude to this deeper alignment of queries and purchases that queues up each creative? We haven’t seen a shred of official documentation on it.
Beyond the black-box Matchdriver solution, Yahoo is the engine that spawned a spending cap technology but can’t shed any light whatsoever on how it’s applied within the system (daily fluctuations are noted from the start, but when spending caps are modified in mid-stream, the pacing mechanism is completely thrown out of whack). The DTC seems to have been developed outside of the actual bid management and ad-serving platform, in that when making a chance you have to make that change on three different levels (account, category and keyword). This draws monumental complications and frustrations in dealing with the DTC. And then there’s the most audacious oversight from poor product design, offering a separation of search and contextual bid markets without the ability to track them separately. Actually there is a tracking solution that would work with website analytics tools in order to separate search and contextual (but it’s rumored not to work that well). And it does not work with your standard set-ups of DART or Atlas
So we’re talking about subtle expansions of well-known products. Those products are well publicized, much better than their implications. We have to get a product manager on the phone to describe those finer points, because our sales reps are largely unaware of them. Let’s face it, Yahoo and Google are the perfect blend of technology and sales, wrapped in a marketing juggernaut. These aren’t necessarily people who’ve used their own products, as would the advertiser. So what we have here is a basic game of show and tell, the engines show us everything but they tell us nothing. With the markets becoming increasingly competitive (possibly due to inflationary practices and product development on the engine side), search marketing has become a game of inches. With that concession, these details matter more and more every day.
An advertiser should be able to take production for granted…if you can’t offer flawless campaign set-up, your product simply doesn’t work. But if you’re aspiring to a higher level of advertiser relations, customer service lies in education. Mirroring a sentiment recently provided by Reprise Media’s Director of Product Development with respect to the eventual MSN launch, MSN has been hiring smart domain experts that approach search marketers with constant questions of how they can do better. Who would have guessed that out of all companies, Microsoft is the one that’s going to get it right?? Because they seem to identify themselves as stewards of client relationships, something that’s somehow gone overlooked in what is currently, and will always be, a service industry.
Bottom line, engines should educate their account executives so they can in turn educate their clients. Because with the breakneck pace of product development we’ll continue to see, the things advertisers need the most is education. Educate us and we can educate our clients. But don’t let it go at the basic product level, you need to help us understand the implications behind these products, the tactical elements, so we can apply them properly. Because otherwise industry PR will continue to frame their search marketing coverage around basic bid management, and that would be the worst PR for our industry.
Randy Schwartz is Director of Strategic Development at Reprise Media
Topics: Google |