23 Musings

ComScore have come out to explain a little more detail behind the release from a few days ago which announced the decline in clicks on Google's paid search ads.

It seems that the reason for the decline in clicks is down to improvements made by Google to increase the click quality (such as only making text clickable instead of the entire ad unit).

One has to wonder whether a letter from Google has triggered this deeper analysis of the issue...

Anyway, with the recent news that click fraud is rising rapidly it's timely for Google to remind us of these improvements. It will be telling to see the Click Forensics data for the 1st quarter 2008.

Interesting take on the role of search marketing in travel here from Elisabeth Osmeloski of Search Engine Watch.

Not completely sure I agree that seasonality will disappear, there will surely always be a place for targeting specific seasonal activities and travel times with certain keywords. Yes you may run the campaign full time but it's usually a good idea to ramp it up in the appropriate seasons.

ComScore have released some data showing that clicks on ads on Google were down 7% in January compared to December and flat year-on-year (actually down 12% qtr-on-qtr).

This is pretty astounding news after the growth Google has seen in ad clicks over the last few years. The thought is that this isn't anything fundamentally to do with Google or any competitor taking market share away, rather analysts seem to think this is a sign of the economic uncertainty we are currently seeing. Times are hard so people click less on ads....

What would be interesting is to see the search volume data alongside this click data to see whether searches have declined or stayed the same.

Needless to say Google's shares have taken a bit of a battering today because of this.

Now I've just been proved totally wrong. I wrote a while ago that paid search budgets may actually start decreasing (slightly) or stagnating by the end of 2007 in travel as marketers got their heads around affiliates, other new behavioral ways of marketing online and moved back to seeing the huge value in SEO. Also a backlash against paying per click has been expected for some time as marketers try to get everything measurable on a CPA basis.

Seems that's not the case though. Robin Goad of Hitwise has a post which talks about some paid and organic search trends in travel and shopping sites through 2007. One really interesting chart shows that the Hitwise Shopping & Classifieds categories paid search activity actually decreased in 2007 compared to 2006. Travel on the other hand grew by 15-20% during the final three months of the year.

Also, there a great chart showing the trend for paid search traffic to the two categories from late 2006 to the end of 2007.Interesting dip around July 2007 there. Anyone hazard a guess at what that may be? I know it's not the busiest month of the year but that's a hefty drop in paid search spend.

Click Forensics has reported on the 4th quarter of 2007's click fraud volumes. They've seen an increase of 2.4% year on year with the latest quarter registering a massive 16.6% fraudulent clicks. That's the overall figure for the industry average according to their index. The figure for content networks such as Google Adsense sits at 28.3% which is huge!

This is really worrying for anyone with a large paid search campaign. I would have expected these figures to drop due to technological advances from the search engines platforms, instead they seem to be climbing.

Google has announced the launch of demographic targeting for Adwords. This will enable you to show your paid search adverts to more targeted customers and thus drive better qualified traffic to your site. Very useful!

This involves site owners sending Google anonymised user data so that Google can then track these people and serve targeted ads. Surely this has to raise questions for data protection? That said, it is immensely useful to any retailer with data on their visitors!

Paid search (according to Marketing Charts) contributed 57.1% of the total spend on online marketing in the UK. The spend was up 44% year on year for the first six months of 2007 compared with 2006.

This just shows how effective it is and where your marketing pounds should be aimed.

Classifieds has shown good growth as well but this is not as targeted so it's obviously not going to get the same amount of interest as search for the moment.

Reasons for using search are cited to be lead generation, driving direct sales and traffic generation. Great to see sales being the equal top reason for using search. For too long it's been seen as a way to drive traffic and acquire leads, when actually it is the most effective way to drive a direct sale available to online marketers.

Tripadvisor have released their figures from a survey called their Traveller Trends Survey. Out of the survey comes some predicitions for 2008 trends that we should all expect to see.

One of the aspects of this (the link above has much more detail) is the TripAdvisor TravelCast. It's a barometer of what’s hot in travel destinations. TripAdvisor engineers have developed a proprietary algorithm that looks at several criteria, including changes in search activity and postings throughout the TripAdvisor site. The TravelCast then predicts the rising stars in travel.

The destinations that are showing as increasing in popularity and expected to be the biggest next year are:

TripAdvisor TravelCast Top Ten World Destinations for 2008
1. Jerba, Tunisia
2. Makandi Bay, Egypt
3. Phangnga, Thailand
4. Kovalam, India
5. Sabaudia, Italy
6. Asilah, Morocco
7. Ko Phangan, Thailand
8. La Plagne, France
9. Yangshuo, China
10. Kotor, Montenegro

TripAdvisor TravelCast Top Ten U.S. Destinations for 2008
1. Sunny Isles Beach, Florida
2. Kitty Hawk (Outer Banks), North Carolina
3. Seward, Alaska
4. Kailua, Hawaii
5. Blue Ridge, Georgia
6. Mount Pocono, Pennsylvania
7. San Marcos, Texas
8. Paso Robles, California
9. Rockport, Texas
10. Copper Mountain, Colorado

Now, this is great insight for online (and offline) travel providers who bid on keywords on search engines. If you are active in pay per click and these destinations match some of your offerings then you should seriously consider looking at the data (clickthroughs, bid positions, bid amounts etc) to see if you can do anything else to get more clicks on these terms. If they are going to be so hot next year it may be worth diverting some of your paid search budget into these destinations over those that don't perform for you currently. Insight like this is highly valuable, especially when it comes from such a big community as TripAdvisor.

Click fraud, the bane of every marketer who deals with paid search, is currently sitting at around 16% of the clicks you may be receiving. The average rate is 16.2% and can be significantly higher on content networks.

The average rate was 13.8% a year ago and 15.8% last quarter so it is rising quite fast:

The average rate of click fraud on content networks such as Google Adsense or Yahoo Publisher Network has risen more sharply:

The graphic below shows a threat map for where click fraud can be generated:

So why this continuing rise? Well as it's coming from content networks I can only guess that it may be being generated by rogue publishers and bloggers who are auto-generating extra clicks on their sites in order to raise their income. This kind of practice could really be affecting your bottom line and giving you a falsely inflated CPA (cost per acquisition) on your paid search campaigns.

Well worth keeping an eye on! And if, like me, you have a large campaign under your remit then contact your agency asap to check that you are receiving any rebates the search engines such as Google provide.

Thanks to Marketing Vox.

Interesting analyst opinion from Bear Stearns on Yahoo here. The analyst who prepared the report put Yahoo as one of their top picks, citing such factors as signs that advertising pressure may be decreasing and initial concerns about Panama (the new Yahoo paid search platform) may have been unfounded.

Most interestingly he cites Yahoo as ripe for acquisition, mentioning Microsoft as a possible suitor. he believes that Microsoft have been actively investigating the possibility of acquiring Yahoo for some time. He also said that he could see any buyout price being at approximately double the current Yahoo share price. Now that has to be tempting for Yahoo!

If this ever happened (which I'm not yet convinced of) it would turn Microsoft into a major player in the online world, instantly buying them a serious amount of advertising real estate and with the potential to compete with Google Adwords in the paid search arena.

Now American Airlines has sued Google for allowing other advertisers to use it's trademarks as keywords in pay per click advertising. The airline accuses Google of selling the right to use American Airlines' trademarks and service marks or "words, phrases, or terms confusingly similar to those marks" to competitors who then direct searchers to their own web sites.

This isn't the first time Google has faced such a lawsuit. Geico sued Google for the same reasons some time ago and lost, and apparently other cases are on the backburner.

This confuses me a little... We use Google extensively for PPC advertising and our highest converting keywords are our brand terms (as you'd expect). Every so often we find a rogue affiliate or competitor bidding on our brand name and we always report this to Google and they remove the offending adverts for us. To enable this kind of response we had to register our brand terms with Google. They don't really police it actively but they do take down offending ads when asked.

So if the above is possible, why don't American Airlines just ask for them to be taken down? I'm guessing that they expect Google to do this automatically and to not even allow the ads to appear in the first place. To enable that would be a hugely complex and time consuming development for Google and a fundamental change to the Adwords system. I'm guessing Google would rather not have to do that. But if American Airlines lose (like Geico) then surely Google should not be taking down our competitors ads (as it's not been deemed illegal)?

Who knows! What I do know is that brand keyword advertising is very lucrative, it returns excellent ROI and is any search marketers meat and drink. Any threat to the way brand term advertising works could have a massive impact on Googles Adwords revenue. If lawsuits like this keep cropping up it is possible Google could ban advertising on trademarked terms for all to stem the tide of subpoenas, that would make PPC a much less attractive proposition!

It seems Yahoo has fundamental problems in their ad sales team (at least that's what I think).

Yet another quarterly profit report that's down on last year and again it's blaming sales of display adverts on the drop. However, everyone is reporting huge investment in online marketing and massive growth in this area. Yahoo have a major piece of internet real estate but just don't seem able to capitalise on it.

I realise a lot of the increased online ad spend is going into paid search with the likes of Google, but display should be Yahoo's meat and drink and shouldn't be an area they are losing money on. Yahoo provides masses of eyeballs for banner adverts but it seems the money is being held back to be spent elsewhere.

The solution? Capitalise on what you've already got! Huge traffic, massive amounts of quality content, huge amounts of page views. Create brand centric areas which can be sold for large amounts of advertising dough and fix Yahoo 360, turn it into a proper social network. Yahoo still has a huge community who use Messenger, Mail, Flickr etc. All these users have a Yahoo profile which could be a 360 profile, fix the social networking aspects (or build them if they don't exist) and you could get some loyalty back which in turn will attract the ad dollars back.

In recent weeks our Adwords campaign has increased in cost a fair bit. This is without any major changes to the campaigns, all of which have been running nicely and costing very similar amounts month-on-month. The conversion rate hasn't dropped much but the cost has increased a lot!

I'm wondering what's causing this? Has competition increased that much in the last couple of months that we should be seeing this kind of increase? Or is our agency missing a trick and not optimising the campaign to their usual effectiveness.

I'm going to deep dive the data over the next week or so to see if anything has gone awry, but I'd appreciate any insights from anyone who has any ideas or experience (especially from the travel industry) of the same?

Google is launching it's Checkout payment solution in the UK at last. It's been in the US for a while and although it hasn't taken over from PayPal it is making good headway. This has been partly helped due to the fact that they are not charging fees for merchants at the moment.

It's an easy to use payment solution which stores your credit card details so you don't have to retype all the time. It's also been integrated with Adwords which is a really good move, imagine click to buy from the search results in the future, that could be a really powerful tool for advertisers!

This from the EFF (Electronic Frontier Foundation):

A Bad Idea From Utah: A Ban on Comparative Advertising
March 30, 2007

The Utah legislature has quietly passed a dangerous law allowing trademark owners to prevent their marks from being used as keywords to generate comparative ads. If this law takes effect, a company like Chevrolet couldn't purchase "sponsored link" space on the Google results page when a user types "Toyota" as part of a search query--at least if the latter term is registered in Utah as an "electronic registration mark."

As Martin Schwimmer notes, Utah's own General Counsel warned the legislature that the law was likely to be found unconstitutional given the burden it would put on interstate commerce. To comply with the law, a search engine that received a search request would have to determine whether a user was located within Utah and, if so, check the search terms against Utah's registry of trademarks to prevent the unlawful triggering of advertising. The cost to search engines would be staggeringly high: "Literally millions of search requests from locations worldwide each day would be subject to verification of location."

Aside from its constitutional flaws, the law is just bad public policy. It undermines the fundamental purpose of trademarks: to improve consumer access to accurate information about goods and services. Trademarks are just shorthand terms that designate the origin of a product. Comparative advertising uses those shorthand terms to provide more information about the trademarked product and competitive products. That's why comparative trademark use is clearly protected under federal trademark law. If it weren't, Pepsi wouldn't be able to tell consumers that more people think Pepsi tastes better than Coke, and Apple wouldn't be able to make fun of Microsoft on national television every night.

The good news is that, given the constitutional problems, the law is likely to be challenged in court. But it's too bad the Utah legislature didn't heed its own counsel's advice and save Utah taxpayers the cost of defending this anti-consumer legislation.

Now, as far as I know Google will stop people from bidding on a trademarked term if they are notified of the fact. I have instigated this myself in the past when we've found someone bidding on our company name. So quite why these lawsuits are happening is beyond me, I know Yahoo etc don't apply the same controls over trademarks but with the bulk of the paid search market poured into Google I question the point of trying to ban it (also find it hard to imagine how it could be policed purely for Utah).

Google is fighting for the right to allow it's paid search advertisers to bid on trademarked keywords. This strange move seems to have come about thanks to Google's ongoing litigation with Rescuecom, the litigation has been ongoing since September 2004.

Google has filed a brief in the case this week which makes a compelling argument as to why sales of trademarked keywords to it's advertisers should be allowed. Google say that companies associate products with competitors all the time in other forms of advertising and that so doing doesn't cause confusion for customers - which is what a trademark is supposed to protect.

Google's lawyers say: "Generic brands are placed next to known brands on store shelves for the express purpose of diverting customers from the brand they are seeking to another, and their manufacturers pay for that placement,advertisers deliberately select magazine ad placements next to articles about their competitors. ... All manner of companies pay for coupon placements selected based on a customer's purchase of their competitors' products. And so on. Of course they are seeking to 'hijack' or 'divert' consumers who have indicated an interest in their competitors' products. That's the point of contextual advertising -- to target ads at consumers who are actively interested in your type of product, rather than indiscriminately at the world at large."

It's a fairly persuasive argument but in my opinion could spell trouble for Google if the rule changed. Trademark terms drive a huge proportion of our paid search traffic and it's the highest quality (and converting) paid search traffic as well (for obvious reasons). Diluting that (which any change would) could make Google a less attractive place to pump our money and make us move to other CPA alternatives even quicker than is already happening.

Yahoo's new paid search advertising system is being offered to new accounts now. Originally it was just offered to exisiting accounts in October.

Yahoo's counting on the new system as a way to grasp some market share back from Google. I believe they need more than just a new system. Simple issues such as preventing people from bidding on tradmark keywords would get Yahoo a lot more love from their advertisers.

However I do feel the main reason Yahoo can't compete with Google on paid search is due to their natural/organic search algorithm being so poor. Users go to search engines to find information, most of that information is found in the natural results so if the natural results are no good..... Without this core audience of searchers the paid results just don't get the volume of clicks and conversions that advertisers want from a pay per click campaign.

Great post on the Central Desktop blog regarding Google's amazing ability to appear top in the paid results for a number of keywords no matter how much other advertisers bid.

I quote: "Google holds the top advertisement (Adword) slot for the following key words:
intranet, spreadsheet, documents, calendar, word processor, email, video, instant messenger, blog, photo sharing, online groups, maps, start page, restaurants, dining, and books (somehow Amazon has managed to appear in the #1 ad slot for 'books').

For spreadsheet, blog and video, in addition to squatting the premium ad position, Google Products also dominate three of the first four search results.

In such cases, Google Product Links and Ads can account for up to 25% of your viewable screen resolution - 30-40% for lower screen resolutions - almost guarantying that users will click on a Google Product over any other search results, sponsored links or text ads.

What this tells me is if you are trying to advertise a product that is competitive to Google, then you'll never be able to receive the Top Ad Position, no matter how much money you bid and spend."

They go on to surmise how much this is pushing up the CPC (cost per click) for other advertisers bidding on these terms and also mentions that Yahoo does not follow the same practice.

It's a valid question, possibly slightly biased by the fact that Central Desktop are a competitor in the online office product world but it's a very good post and worth a read!

Personally I don't feel this is monopolisation, more a savvy move by Google (who wouldn't take the opportunity to drive more business to yourself) but I can see this being blogged about for quite some time.

At the time of writing this Google is no longer appearing in paid results for some of these terms, but thought this worth of comment anyway...

So we're in the web2.0 age and the web's advancing quicker than ever, new interfaces, social aspects, interaction of users with data etc etc. One thing that hasn't changed is the model us advertisers pay for things like price comparison websites, pay-per-click. Google has used pay-per-click for it's sponsored listings Adwords product sucessfully and I wouldn't imagine that will change, but for your average price comparison website which is supposed to deliver much more qualified users isn't it about time they moved to a cost-per-acquisition model?

Being in travel we use price comparison sites a lot and derive a lot of traffic from them although not a huge amount of sales. Moving to a CPA model will complement the work these sites are doing to integrate content into their offerings, they're trying to increase the quality of the visitors they send through to us and so should increase the conversion rates. Running on a CPA model could then prove to be in their interests as they may drive less clicks ultimately but with greater conversion. Staying on a PPC model would mean they will lose revenue rather than gain from all their hard work to optimise the listings.

With cookie tracking so easy to implement and fairly accurate, keeping a track of the sales isn't a problem and most people do this anyway. Moving to CPA will also remove any risks of click fraud as there will be nothing to be gained anymore.

As I said, I'm not sure Google would ever go down this route for Adwords, search is much more about driving high volumes of traffic and the visits are never going to be as qualified as those from a price comparison site. PPC seems a little old fashioned as a business model for a website that should be driving sales rather than purely traffic!

Clickriver is now in public beta. This is quite an exciting development from Amazon/A9 as they've chosen to build their own system to do this rather than syndicate Google Adsense or another solution.

The reason for going it alone is most likely because they really don't want any competing products displaying on their site. They're looking for complimentary advertising rather than competitive.

It will be interesting to see quite how much functionality there is in Clickriver. As a travel company based in the UK we'd want to only display ads on amazon.co.uk for example. We'd also only want our ads to appear when someone mentioned a destination in their search or were looking specifically at travel related products. So having this kind of functionality available to you through whatever interface Clickriver uses is really important.

As an official paid search fiend I'm quite excited by the prospects of this and it leaves scope for other companies to do similar. Tesco.com perhaps are big enough to do something like this.