• Chas

    How much stock do you own in Google, or were you paid directly for this piece?
    Thankfully, I have found an alternative that actually serves the customers that provide their meal-ticket, instead of giving them 40 lashes, and it is out of the dictator, Goog’s reach.

  • http://www.stormstudios.co.uk Simon

    Will Google just be grading landing pages for pay-per-click web sites, or will ALL web sites be graded?

  • http://www.rimmkaufman.com George Michie

    Chas, no question that this is just more bad news for scam artists. I’m okay with that.

    Simon, great question. The heart of what Google tries to do is to match user searches to the most relevant pages available, so pretty clearly ‘yes’. The Panda updates have been about better identifying low quality pages and on the flip side, high quality pages.

    In some ways I think they’ve assumed that advertisers would pick good pages for themselves for all the reasons I outlined, but have found that in many instances they don’t. Ane, they’ve recognized the second-generation of AdSense Spam sites produce a lousy user experience and taint the sponsored links as a class. They’re trying to eliminate that threat to their business as well.

  • http://www.findmefaster.com Matt Van Wagner

    George – good analysis, as always.

    Sure hope Google gives much more transparency on LP QS than it does on KW QS.

    I’ve got many QS4 KWs that are my best converting KWs by far, and yet, because Google ties QS to the economics of the auction, I am likelye paying much more than I should be with a QS of 4. Clearly users view our pages as relevant or else they would not do business, so if Google’s wisdom of the crowd were always right, these KWs should be 10s.

    Now, as for landing pages. If Google are really serious about wanting to improve the quality of websites/landing pages in the eco-system, then Google should provide much greater transparency in how they develop their LP QS.

    Google should trust us that we will try to game the auction if we know too much. What could possibly be bad about knowing exactly – exactly – why a LP is considered to be sub-par (or super-par, for that matter)

    Microsoft’s AdCenter’s approach to QS for LP is to give as much info to the advertiser as it can – aka transparency – without punishment by tying QS into their economic model in the auction. I prefer that approach because it implies equality and trust between them and their advertisers.

    How can Google possibly determine what LPs will work without applying conversion rate data?
    Conversion rate (CVR) is to LPs what CTR is to ads. CTR is clearly king on ad relevance for Google’s auction model. Why wouldn’t Google want to apply the same ‘wisdom of the crowd” calcluation apply to LPs?

    How can Google possibly avoid the temptation to use, even in a generalized sense, conversion rates of page types in modelling best practices? How can a best practice for a LP NOT include conversion rate as a strong criterion?

    In general, I am in favor of good advice from trusted, respected business partners, even if delivered in heavy-handed, data-driven and costly ways like Google’s QS does for ad auctions.

    I just hope that Google will find a way to be much more open and enlightening about what it likes or dislikes about a particular page.

  • http://www.findmefaster.com Matt Van Wagner

    Oops, I meant to say, “..Google should trust us to NOT game the auction…” in my comment above…

  • http://www.rimmkaufman.com George Michie

    Thanks for the excellent comments, Matt. I agree with you that conversion rates are the ultimate metric for determining the quality of the user experience in eCommerce and for lead gen folks. It’s a bit trickier for publishers and information sites that make money off of page views and time on site. Moreover, Google simply won’t have conversion data for many of its advertisers who won’t share that info. Giving preferential treatment to folks who share conversion information might open up a whole can of FTC issues.

    Interesting to see how this plays out, but if my read is right folks like you and me are already taking care of this to the degree that Google expects us to. This may just serve to help those who manage campaigns well and punish the sloppy and slothful.

  • http://about.me/alexedlund Alex Edlund

    Great analysis George.

    To Matt’s point, it seems to me that if Google is really trying to assign a quality score of a landing page, they need to include metrics that matter. Bounce rate is one of the best metrics in determining relevancy and user engagement. Still, I don’t envy them for taking on this task.

  • http://ReachableHQ.com Charles

    This is great for anyone who has already been using well crafted landing pages. The key here is “Does this page match the user’s intent?”. Google is trying to do this as best as possible, when it does it is good for Google, good for the user (the searcher) and good for you (the site owner).

    Even starting with a page title that closely matches the PPC ad’s promise will help the conversion rate. We have talked about this before on our site (http://reachablehq.com/blog/how-to-choose-a-landing-page-title/). — ‘Google’s guidelines are that “Users should be able to easily find what your ad promises”. A page title that closely matches your ad text is a great place to start.’

  • http://jamesbutler.net James

    Mr. Mitchie,

    Your comment: “…they’ve recognized the second-generation of AdSense Spam sites produce a lousy user experience and taint the sponsored links as a class. They’re trying to eliminate that threat to their business as well.”

    If only that were true … but it isn’t, because AdSense spam sites are big money-makers for Google. In fact, just the opposite is true. (http://adwords.google.com/support/aw/bin/answer.py?hl=en&answer=50002) AdWords reps even argue that clicks from “parked domains” are MORE valuable than clicks from Google Search, because the “parked domains” are so single-topic focused. (Totally unsupported, in my lengthy experience.)

    The real problem is that Google heavily values AdSense spam sites (a.k.a. “parked domains”).

    Google, bless them, is simply using spin terminology (“good user experience”) to try and communicate one of the challenges they face as a somewhat-indiscriminate advertising vendor: What do you do when a substantial percentage of your customers (advertisers) simply game your system?

    The first answer is … make them part of your Network! Instant validation! Problem gone!

    With this latest pronouncement, Google says that the landing pages should be “higher quality”, implying or else they will punish the advertiser by (a) making their ad less visible and/or (b) making it more expensive for the advertiser to participate.

    And that’s Google’s prerogative as an ad vendor, for sure, but if they are serious, then “parked domains” should be entirely out of their Network. Until that happens, all of their statements about “quality” need to be taken with a big grain of salt. “Do as I do” needs to come into play, somewhere, or else it’s all just weaseling.

  • http://www.rimmkaufman.com George Michie

    Alex, Charles, thanks for your comments.

    James, your point is very interesting. I do think Google struggles to algorithmically differentiate between quality sites and well-crafted spam. That said, I really do believe they have enough vision to want to protect the user experience. If customers sour on sponsored links it’s game over for Google. The whole reason Google created QS and its original minimalist landing page component was to eradicate AdSense Spam pages, and it largely worked. The well disguised pages survived though as you point out.

    The auction dynamic is complex. The fact that spammers are big spenders on AdWords doesn’t mean Google would lose much money by crushing them. Other higher quality listings move up the page and take that traffic, so Google’s loss is ‘only’ the delta on the CPCs, not the whole amount.