I’m not going to rehash the background and deal info, referenced at SearchEngineLand, but instead, I’ll try to explain what I see the market impact as being.

I’ve long argued the point that we should be able to price keyword by partner on all the major search engines, as they all have quite different conversion rates and “Not every click is created equally”, to paraphrase. The AOL deal has forced Google to make this move, and I think it will probably hurt their margin numbers and not just their revenues. Overall, it’s probably close to a zero sum gain as the money will flow to AOL, but their cut will just reduce. I haven’t done the math, but they might be just slightly worse off on revenues, but definitely down on margin.

Also, this impact will hurt many of Google’s partner sites that relied on high conversions from AOL to “slip” clicks through the door at higher than market value (respective earning or value per click), because of the fact that marketers cannot price each partner differently.

This is quite a complicated post, so please bear with me – I’ll try to explain this as simply as possible.

Currently, when buying Adwords ads, you pay a single price for a click & keyword that goes to Google’s distribution network.

For example (and I’m using hypothetical, BUT REALISTIC numbers here), let’s say that a keyword received 100 clicks and the breakdown by partner is as follows:

25 Clicks from AOL – 5 Conversions
50 Clicks from Google – 4 Conversions
25 Clicks from Google Partners – 1 Conversion

Total Clicks = 100 x CPC of 25c = $25 in costs
Total Conversions = 10 divided by $25 in costs, leaves us with a $2.50 CPA (Cost Per Acquisition)
Assume an average position of 3 and as you can see from above, AOL has the better conversion rate.

Let’s assume that your breakeven point is $2.50 CPA. Now, because Google runs a single blind marketplace, there is no price discrimination, which means that there is cross subsidization of keyword value (you pay an average price for everything, yet some clicks are worth more and others are worth less).

For the record, our logs indicate that AOL has the best conversion rate across all Google partners, and this simple means that if there is a separate marketplace for AOL, that prices across the Google Adwords network (especially direct) should fall, IN THEORY (see conclusions).

Now, let’s assume that we removed the AOL numbers:

50 Clicks from Google – 4 Conversions
25 Clicks from Google Partners – 1 Conversion
25c CPC x 75 Clicks / 5 conversions = $3.75 CPA

By AOL moving out of the marketplace, this would effectively mean that the remaining traffic would become too expensive, because AOL would not be cross subsidizing it (and obviously, they know this already, and that’s why they want their own marketplace). This would mean that prices would drop in order to offset the drop in traffic quality.

What would this mean for AOL?

$2.50 CPA x 5 Conversions = $12.50 / 25 Clicks = 50c CPC – almost DOUBLE what Google was raking in for them by cross subsidizing, is what the merchant would be willing to pay direct on AOL, because of the higher conversion rate.

Now, here are the possible conclusions vis a vis Google CPC’s:

1. Google CPC’s will drop as marketers move their spend directly onto AOL and adjust ROI’s according on Google. This will impact Google’s financials, as they will have to pay out a lot more money to AOL.

2. Google CPC’s will increase as dumb money floods the market because marketers are either too lazy or overworked to load AOL campaigns (highly unlikely in the long term) and instead they make it more profitable for Google to show Adwords direct ads for AOL, than AOL marketplace ads. I doubt this would occur though, but it’s a possibility.

3. Google CPC remain unchanged as AOL take-up rates are too low to impact the overall conversion number in the short to medium term.

So, logically speaking, by removing the “bane of mankind” *according to me* (Cross Subsidization) from the Adwords system, we as marketers will be able to align ourselves far more closely with the value per click.

Where do I get this insight from? Well, as an affiliate marketer, we have to watch the earnings per click very closely. EPC is a key metric and we are also able to distinguish (using our proprietary technology) the different conversion rates and therefore effective EPC’s per network partner for all the search engines. This data allows us to derive very accurate statistical models and also understand how the market is interpreted (well, as best as possible) by other players.

Every cent we spend is our own, so we have to ensure that we’re as closely aligned to value per click or EPC as possible, and therefore we’re very excited about the AOL prospects, but realise equally that it’s a lot of work.

I think this is a great move by Google and I applaud them. They’ve thrown the gauntlet down to the other search engines, and let’s see how they respond!

How does Impression Share reflect on my campaign’s performance?

Impression share is a metric available at the campaign and account level for search.

* Campaign: If an advertiser has more than one campaign competing in the same target market, the resulting impression share for any one campaign won’t be penalized if it loses an impression opportunity to another campaign from within the same account.
For example, let’s say Campaign A and Campaign B both belong in the same account. Out of the ten potential impressions, Campaign A shows twice, and Campaign B shows five times. A competitor takes the remaining three impressions . A campaign-level report will show an impression share of 40% for Campaign A (2/(2+3)) and 63% for Campaign B (5/(5+3)).
You can see that the impressions Campaign A lost to Campaign B is not factored in to the denominator when we calculate Campaign A’s impression share. This example illustrates that it is a lost opportunity for your campaign only when the competitor wins the remaining potential impressions.
* Account: Reports at this level will not penalize your impression share from the competition across all your campaigns.
Let’s take the same example from above. An account-level report would show an impression share of 70% ((2+5)/(7+3)) for the entire account. If you look at the denominator, you can see we factor in only the impressions you’ve won across all your campaigns (Campaign A and Campaign B) plus the opportunities that were lost to outside competition.

This will (with further understanding) certainly help clarify if Search Affiliates do better job than Search Agencies. There has always been uncertainty from the merchant side, with this data – there can definitely be closer relationships between both Search Affiliates & Search Agencies (or worse, who knows!).

My belief is that through the use of Search Affiliates (direct linking), we will now be able to quantify the benefit that we offer to a larger range of merchants, than our existing clients. The ultimate aim for any merchant will be to get 100% impression share, and given the sheer volumes of keywords and variables (bids x keywords x positions and then some) that are out there, I don’t think it’s possible for just a single company to be able to achieve this. Search Affiliates will be an even more vital part of the Search Marketing ecosystem going forward!

Kudos to Google on this move and providing the market with more information – I think it’s a very positive move especially for Search Affiliates!

For those of you who do NOT read the comments section, IT WAS A JOKE! Google is NOT dropping CPC!

Other pranks by Google were these ones.

With this change, Google will also be showing image ads on SERP (Search Engine Results Pages), in the form of a 234×60 banner, after the search box. These image ads will start at a $100 CPM rate and bidders can participate from 1 April 2007 (today).

Yahoo have responded to this by announcing that they will continue their CPC policy and remove image ads from the My Yahoo homepage in an effort to promote their search box more prominently.

UPDATE (3 April 2007): THIS WAS AN APRIL FOOLS JOKE – DO NOT TAKE IT SERIOUSLY

Basically, James already details what I said during the show, but just to clarify what I mean by Return on Effort, here is my short and simple take on things:

The question was asked as to whether or not it is worth spending time on 2nd tier (non-Google/Yahoo/Ask/MSN) search engines and running campaigns with them.

There are a couple of key issues here, one is market growth (momentum) that Google in particular has, and the second is ROE (Return on Effort)
Let’s make the following assumptions for the US market search engine market (taken from VastPlanet):

Google Market Share = 53% (with AOL)
Yahoo Market Share = 28.1%

MSN Market Share = 10.5%

Ask Market Share = 5%

The Total for The Titans is a whopping 96.6%.

Now, until Snap, Become, Miva & all the other 2nd tier engines send traffic out of the massive :- combined 3.4% market share that they have & I can’t see the logic in advertising with them, and here is why:

If in one of our campaigns at Clicks2Customers, we have to allocate a campaign management resource to setup a campaign on a 2nd tier engine (which we don’t deal with). Now let’s assume for a decent sized researched campaign of 5,000 keywords with dedicated ad copy (as all engines are different and have differing editorial rules), it takes them 100 hours to do (and that’s quick, using our technology and existing processes).

Let’s assume that currently on Google, we are running with 50,000 keywords and generating $100k a month in sales and in our category, we’re getting 5.3m impressions (searches) per month. All things being equal, by the ratios above, the maximum searches we would get out of ALL the 2nd tier engines combined, would be 340,000 searches with 50,000 keywords. Let’s further assume that we went with the largest 2nd tier engine (not even sure who that is) and the engine had a 20% market share, then I’m going after a market of 68,000 searches related to my product/service – if you divide that further with the fact that you’re only loading 5,000 keywords – it would get scary, so I will neglect to include this in my calculations.

Again, ceteras paribus, if you just work out the back of the envelope stuff, then the absolute maximum that this traffic is worth to me if it converted even just as well as Google does is $1,283 in sales (which I highly doubt, as there are large amounts of Clicks Fraud on 2nd tiers). And that’s with a 20% market share which is not even possible in such a fragment tail-end market.

So, assuming I could spend my 100 hours on Google, and push my campaign performance up by just 10% with an extra 5,000 keywords, then I would be pushing the needle on my revenues by $10,000 (a nearly 400% increase in ROE), why would I bother with 2nd tier engines? I know the argument (from the 2nd tiers) is that it’s cheaper, etc – but at the end of the day, spending those hours improving clickthrough rates and other metrics like conversions etc, will translate into greater savings anyways on the majors like Google – so I don’t buy that argument.

Most Google PPC campaigns I have seen are not even 50% at peaking in terms of digging into traffic in the long tail, and most people are so worried being on other engines that don’t matter and waste their time there. Mine the Google keyword gold instead, and when you’re finished making triple digit gains, then go visit Yahoo and then MSN, and then finally, Ask.

Lather, rinse & repeat.
Someone today said to me that no one ever made a fortune by worrying about the numbers after the decimal. I think that this definitely holds true in this case.

Hi Vinny,

I have just recently found your blog, and greatly appreciate your views on search marketing, with a SA twist!

I have a question that no one has been able to answer and I hope you can help?
If I am based in the UK, but which to target another country only, such as
SA or USA, what would be the best option in terms of currency to use?
When I would create an adwords account I would select “based in UK”, but
then under “I wish to pay in” I can select the local currency.
Does the option I choose place any advantage or disadvantage over my cpc or ad position?

Your help is very much appreciated.
thanks again,
Marc Samouilhan

Thanks for the question Marc.  Funnily enough, we dealt with this issue many years ago, and I never followed up to see if it changed – maybe other readers can give insight on that.  When we posted this question to Google, the response was that all the bidding in the background is done in US Dollars, and then converted to the currency of the advertiser, using a daily spot rate.  In theory then, all bid increments should be roughly equivalent to increments of USD (cents) x Your Currency conversion rate.
Based on that information that we understood to be true at the time, I’ll answer your questions.

1.  The best currency to us, is US dollars.

2.  If you’re running multiple campaigns for multiple clients, it makes sense to try and standardize it to avoid currency nightmares due to fluctuations.

3.   In theory, there is a disadvantage during your campaign price analysis, as you now have to factor in currency fluctuations into the bid price increase/decrease.  You can also overshoot or undershoot the market CPC by placing a bid which a high currency variance (i.e. paying 5p instead of 9c, because the next step up charge is in 2c increments due to the strength of the pound).

My recommendation would be to use USD, based on the information I got a few years ago.  Does anyone know if this has changed?

Our show starts at 10am EST (5pm South African time) on Thursday March 22 and it promises to be really worthwhile.  You can signup at eComXpo for free and the rest of the shows started earlier today and continues until Thursday.

On the note of Affiliate Arbitrage, Chris from SearchAnyway.com  covered Brian Geddes piece on Search Arbitrage.  There is often a misconception due to non-standard terminology in the industry.  Let me clarify:

Both Search & Affiliate Arbitrage can be construed as affiliates bidding on search keywords and sending traffic directly to a merchant in exchange for a Cost Per Action payout, whereby the affiliate is paying the search engine on a CPC basis.  Brian’s definition of Search Arbitrage is when scummy marketers use landing pages full of adverts being paid to them on a CPC basis, and buy CPC traffic from the search engines, on the basis that they should earn more per click than the click costs them on the search engines.  To me, this is commonly known as Click-Pimping, something that I do not consider good for either the advertiser or the end user/searcher.  I really call on the search engines to take these “pimps”  to task and remove sites that offer no value at all.

The key difference is that CPA-CPC arbitrage has implied risk reward ratios, and the more efficiently the campaign is managed, the better for both merchant and marketer. CPC to CPA arbitrage also means in general that the arbitrager is using the company’s display URL in order to increase reach across the long tail of keywords in order to generate cost efficient click prices.  Google is already gearing up to do this now, indicating the value in this space, as per my previous post.
CPC-CPC arbitrage has no value to anyone, except the arbitrager – the end user is irritated but doesn’t realize that they are generating money for the “pimp”.   The additional landing page is of no value.
Regardless, we are left with a misnomer around what Clicks2Customers (CPC-CPA arbitrage) does, and what “Click-Pimpers” do.

My suggestions for names for CPA-CPC arbitrage are:

• Performance Search Marketing
• Affiliate Search Arbitrage
• Search Arbitrage
• Affiliate Arbitrage

With regard to CPC-CPC arbitragers – let’s just call them Click-Pimples and be done with it!

Search Engine Watch believes that Commission Junction’s days are numbered – I don’t think so! Andy Beal also has a similar view. We’ve been running CPA campaigns through Google for nearly 4 years now, and I think Google vastly under-estimates the risks and relationships at play with CPA marketing. The biggest concern though, is that Google’s internal arbitrage of CPC to CPA (which is what they’re doing, effectively), pushes prices CPC prices up in the short term, while they make mistakes that we’ve forgotten how to make, in our Clicks2Customers business.

Also, from the Inside Adwords blog, it’s not clear how they will deal with chargebacks – I’m guessing that the merchants have to factor this in? Can you imagine what’s next? Click-Order-Return (COR) fraud (i.e. Website owner clicks a merchant, places an order – merchant pays Google, Website owner returns goods – Google doesn’t refund merchant and Google pays Adsense site share of CPA). What if I’m a Google stockholder and I make a $1m purchase in order to boost the earnings, and then return or cancel the order – in theory, Google still gets paid and their stock goes up, but the merchant is out of business – just summising here, but I still don’t think CPA is viable for Google.

As I said in my previous post on this topic – Google has a smart bunch of guys, and I’m sure they will figure it out!

So the big question:  Is Google using Adwords CTR & Performance data to influence their natural listings? We believe so!