If you don’t own the sale, make darn sure those who do need you.

Google’s PPA has launched a litany of articles ranging from an acknowledgement that this is just one more arrow in their quiver, to “Oh no, they are going to put affiliate networks out of business,” and even “PPC is going away.”

For me, it is just another arrow. Unlike PPC, where there were a few so-so players (with many more not so good ones), and they could dominate just by being very good, the PPA world has competition that is quite competent; just as radio sales, t.v. sales and newspapers.  But, this announcement brought to mind the broader guessing game: What’s Google up to?

When the company launched into radio, I though it was far a field. Ditto for t.v.; newspapers, not so much, but still out there. My main contention was and still is that agencies won’t simply step aside as Google develops direct relationships with advertisers, or forces blind buys to media. Advertisers themselves are not necessarily interested in disintermediation either. Well, my perspective was skewed by the nature of my interaction at Google, with the sales and agency relationship groups. While still valid (I believe), I don’t know that it is the relevant perspective.

Google is fundamentally a technology company. The sales side has benefited greatly from the door-opening-great-product, relevancy based PPC Search. They have always said their goal was to make the advertising buying process more efficient. But it was always communicated as something over which Google will have control in the relationship. Media placement decided by Google and the goal of developing direct to advertiser relationships have given me pause. Efficiency is important, relationship control is paramount.

Well, most of the transactions are going to take place with or without Google (SMEs are a different and very viable prospect). But, Google can make them better. By interjecting themselves at all levels, Google can learn what makes the process inefficient at every stage of the marketing communications chain (brand / product awareness down to acquisition), and then offer up the tools to improve it. They become the conduit through which all processes are connected and managed.

Ultimately, I do not see Google owning the advertiser / agency – to– media relationship. But, they may just build the tools on which the agency – media relationship depends. They become the Intel inside the ‘machine.’ Agencies can claim a level of proficiency and acumen as they sell their use of the “Google Media Engine” as part of the tactical implementation of their unique agency strategy skills.

When we sell our skills as agencies, we focus on the strategic abilities we have. Uncovering the unseen, identifying opportunities and ultimately growing our client’s business along side them as no other agency can. Too often though, we get stuck on our media buying prowess. It boils down to efficiency – how much do we get for each dollar spent.

The truth is, any good agency, and there are a lot of them, can, or has access to efficient media buying. It is the minimum price for entry. Nonetheless, we spend a fair amount of time convincing prospects we do this better than anyone. It is a commodity. The real value is in identifying the strategic value of a media and being able to integrate it into the overall objectives.

If Google’s tools can abate this question somewhat and allow us to focus on our truly unique strategic abilities, then they will add a lot to the equation. If, however, they simply try to interject themselves between advertiser / agency and media, I don’t believe it will work out for anyone.

Embrace Risk. Then optimize it

Compensation models vary for interactive Marketing services. When they are negotiated, both sides want to shift as much risk as reasonable to the other side. This is why agencies like the commission model. No matter what happens, they get their check. Advertisers like performance models. They only pay for what they get, even if they get less due to factors outside the scope of advertising (I’ve been hit with distribution issues in the CPG industry). Typically, these two meet in the middle.

One such model is at the heart of the online performance marketing industry, specifically "cost per." It may be an action, lead, customer or other quantifiable metric that can be sourced to specific efforts.   

At the resent IAB Leadership conference on performance marketing in Chicago there was much discussion on the issue of the most common method: paying the same amount for each like lead (meaning amount of information submitted.) In general the conversation focused on moving to a pay per media source, allowing the cost and value of the source to be evident to the client. So, Instead of being paid $50 per,   the client will pay based on the relative cost of the Media. The idea behind this is transparency.  If one media cost $25 and the other cost $50 then the agency is paid $30 and $60 respectively with a 20% commission. Or its cost plus. The media cost, plus $5 per lead, not exceed $50 per lead net for example.  

Before I get to my perspective, a little background: The company I work for, Leapfrog Online is a customer acquisition company. We do not get paid until our partners get a paying customer. We front the media cost, so if something doesn't work, it comes out of our pocket.  We operate a closed loop process of integrating our experience and backend with our partners’ systems in order to control the media and the experience. We track everything to the value of the customer it generated. I bring this up for 2 reasons: 1) like other quality online firms we track EVERYTHING (I think even more so, but that might just be pride talking) and 2) more so than other firms, our financial success depends on the quality of the lead and our ability to optimize the quality by controlling and adjusting the media, messaging and experience based on the lead source (because we don’t get paid until the lead becomes a sale.)           

Okay, so what does all this shameless self promotion have to do with the IAB Conference on lead Gen? A lot. The real issue is not the quality of the media source. Rather, it is the ability to optimize during the entire sales process. If a client treats every lead the same, then the cost of the media is the only variable that can be used to equalize per customer value. When this happens, the obvious next step for the compensation discussion is limited to lead source, or media.

If an agency is on cost plus, or percent of spend model, they are protected. Therefore, the client needs to be vigilant in managing the media sources, because the client absorbs all of the risk (caveat – the agency will be fired if the leads are junk). The challenge here is that clients are typically risk averse. According to a February 2007 ARF study, 10% of companies budget $0.00 for new initiatives or higher risk media. Another 74% have 1%-25% set a side (I’m guessing more are closer to 1% than 25%.) I’ll get more into this as an issue a bit later.

If, however we work with an average compensation per lead, or sale, the client is protected and the agency takes the risk. This is good for everyone, if the client and the agency can mutually track back to quality, and optimize accordingly.

Consider the following media based compensation scenario with a client target cost of $50 per unit (lead, sale, trial, whatever counts as a unit):

If we work on a media level compensation, media 2 will not make it beyond the first month (likely not beyond the first week). On the surface, it appears inefficient and the client will pay $28,000 more than they have targeted. This is 1,000 units that will not make it back into the plan.

At this point, I have heard people say, “at that price, it shouldn’t be in the mix.” This is the typical risk-averse perspective. If a client is operating under monthly and quarterly financials, a negative $28k is going to stick out like a sore thumb. On top of that, the agency still gets well compensated even though the units are at a loss.

Now consider the opportunity perspective if the agency has the risk:

In month one, we take a bath on media 2. But, we are the experts. If we control the media, messaging and experience, we can optimize it. By month 2, we lower our losses, and our percentage compensation is good again. However, the real value comes in month three. With the increased efficiency, we can roll new media into the mix and start the process over again.

In the first scenario, we can make more money in month 1. However, we lose opportunity down the line. The client also loses. They give up a potential 1,000 additional units per month and the opportunity to gain quality volume from additional media. Even if we find that media 2 can only support 500 units at an acceptable price, then that is 500 more than we would likely contribute under the first scenario.

At the heart of this are two issues: Control and Trust.

Control over the experience is paramount. Either direct agency control, or significant agency / client partnerships providing the agency with influence and access to the client’s experience. Effective control can not happen without very detailed tracking and analytics. Absent these, all bets are off.

Trust is at the heart of every partnership. Sometimes our reluctance to offer information results less from a lack of trust and has more to do with technical limitations. However, we must have transparency in both directions to make this work. Agencies must know the complete path (online and off line) to customer acquisition in order to truly understand and optimize the customer experience. While I am vehemently against micro managing of the online programs by clients or client service teams, agencies must be willing to open up on where they are placing the advertising and the relative value (easily enough done with proper tracking).

Every partnership is different for various reasons. But, to make the best of it, we need to stop seeing risk as something to be avoided, and start seeing it as something to be optimized.

AOL Search Back, a second chance at the click you missed the first time

When it comes to search, it used to be that you bought the keyword, it was clicked or it wasn’t; if it was, the person bought or he didn’t.

Then we saw “re-targeting” from companies like Revenue Science and a few others. If a person clicks, and he doesn’t buy, you show them ads (it was just text, now it is text or display) throughout the Revenue Science network of partner sites. You know these folks are interested, but you didn’t hit the mark the on the first click. Now you can try different messages and keep talking to them until they convert, or for a set period of time.

But what if you missed the mark on the click? Almondnet developed technology to target people based on the searches they conducted even if they did not click on your ad. The problem is, they are focused on verticals (mostly tech). Not something most mainstream advertiser can leverage. Around since 1998, they have learned a great deal. They just need to scale.

Well, this week at the AOL Road Show in Chicago, AOL mentioned (rather casually) “Search Back”. Partnering with Revenue Science on BT, and leveraging their search volume as well as there own network of sites, AOL takes post search (not just post click) targeting to mainstream advertisers. There properties cover approximately 90% of US internet users. By knowing your audience, you can message well targeted users even if they did not click on your search ad (though, if they converted with someone else, this is too late). If you really understand them, how about targeting them just before they are in market (think search on “prenatal care”, and mini van ads in about 9 months).

I was critical of Google’s BT search approach (personalized search) because we could not easily segment messaging (yet). AOL has addressed that. With Revenue Science’s optimization technology, this is going to be leaps above where we are.

The rumor is that we should see something in Q2. There are categories that can be targets, or you can have custom targeting (about 200 keywords). AOL was presenting this as part of their overall BT package in February. The timeline appears to be consistent.

Click Fraud, a "problem" advertisers do nothing about.

It was a curious set of arguments that followed Google’s release of .02% click fraud. Most centered a round the concept of “we don’t know what we don’t know.” Therefore in our perpetual state of ignorance, we should expect that Google solve all our problems. We have gone from wanting them to stop click fraud, to wanting them to make up for technical issues SEMClubhouse.

What I found most intriguing was the argument that since the .02% click fraud rate is based on the advertisers who have actually lodged a complaint , then it can not possibly accurately reflect the true state of click fraud. Why? It is because the vast majority of advertisers do not lodge complaints. In fact, they do not even measure it.

(Donna Bogatin does a good job of discussing the soundness of the number itself and Gord Hotchkiss on why he believes the Google Number here)

Think about it. Click fraud is an overwhelming problem. The evidence to back this up: Since the vast majority of companies do not measure click fraud, then they must be victims of it.

So lets get this straight. Google says they block most invalid clicks – ranging up to 10%. Of those they do not block, but are brought to their attention by advertisers, they find .02% is what made it past their detectors, and was actually billed to the advertisers (See Donna’s discussion, as this may be .02% of all Google generated clicks, not .02% of the clicks for the audited advertisers – big difference).

What the antagonists are arguing in this case is that the real click fraud is not present with those who lodge complaints, but is actually taking place with those who are silent.

Well, here is my good old capitalist perspective. If a business is not concerned enough to take the steps to detect and report click fraud on its own, then we should not force the search engines to baby sit them. If they bring it to the search engines’ attention, then they should receive a fair and vigorous response designed to make them whole.

The bottom line is that (despite the anti-Google block’s contention), click fraud is bad business for Google. As soon as the metrics are not profitable, companies will stop and the engines will lose. Keeping the issue in check is in their interest.

However, perpetuating panic is in the interests of ‘click-fraud-detection’ companies. Keep in mind that these companies measure the clicks you received that should not have been billed (Shuman Ghosemajumder addresses some of this). What they are not doing is identifying the click for which you were billed, and segmenting them out for reconciliation. When the companies presented the findings to Google, the vast majority of the “fraudulent” clicks were never billed – things like quick multiple clicks, back button clicks, etc.(see Google’s explanation here)  If all you look at is the log file, then you might well believe that you are a victim. But, when you actually count the clicks for which you were billed, the argument falls apart based on Google’s response.

What I have not seen is a counter response to a Google reviews. I have not seen a company show that they were billed for 100,000 clicks, but their log file clearly shows that there were only 80,000 valid clicks, and Google ignored it ( I have heard some say they’ve gotten 10s of thousands of dollars in credits from Google, but no proof). If you have seen it, please provide the links.

Yahoo! Panama, the best is yet to come...how you use it

The new search algorithm has been released for about a month now. I have seen several ‘out-of-the-gate’ measurements from agencies and measurement companies alike. Click rate is gone up, CPCs have gone up, or down, conversions have changed by X…

I think all these are very interesting as a metric of how our old ways to manage a Yahoo! account played out in the new system. But, the real impact here will be seen over the next few months. We will start to leverage Panama when everyone starts to rotate messaging and measure CTR by ad and carry this through to conversion rates. The real performance of Panama will be seen when we drive it around the town a few times.

Certainly, we will see improvements in performance. Not just because Yahoo! will calculate relevancy, but we will be able to adjust ad copy, simultaneously rotate multiple versions and compare the ultimate ROI by ad. Yahoo! will be able to measure the CTR, but we will be able to measure the conversion. As most of us look to the conversion, we would gladly give up non-productive clicks while holding our conversions steady, or increase them. So, our CTR goes down (but so does our costs), while our conversions go up. These will be an interesting next couple of months as we optimize the campaigns rather than just compare them to the old system.

Google Click Fraud Number...so what will you do?

Okay, now the world can move forward. We finally have a number from Google on Click Fraud (0.02%). Now, ask yourself what will you do differently? What if they said 5%, or 10%? If you answer anything other than “nothing”, then why have you not made these improvements before?

For me, click fraud amounted to two things:
1) potential cost recovery
2) an ego hit because some schmuck out there got the better of me

As for my campaigns, it means very little. If you do search right, measure to the margin per transaction, then the click fraud issue is already accounted for in your metrics. You’re already optimizing your programs, so there should be no reaction to the click fraud announcement.

It’s not that we should ignore the problem. By staying vigilant, we prevent it from ballooning on us and turning a once profitable medium into a loss, or at least marginalizing the profit. However, by staying focused, we can mitigate the impact of click fraud, and improve program performance.

Giving up control is good for Quigo, and the rest

It is the number one place to work in America according to Fortune Magazine and at the heart of the reason is that it is all about the people; the atmosphere, the food, the events. Google knows how to treat the employees. They culturally understand the dynamic of interpersonal relationships in making people happy.

This understanding expands to the way they treat their advertisers. At Google, I have worked with some of the nicest people I’ve met. They cultivate the relationships, understand the importance of good communication and go out of their way to make us feel important within the scope of business.

However, I wonder how Google views this dynamic between businesses and people outside its scope. They build tools that interject themselves between advertisers and publishers, agencies and clients. I looked at this before when it came to radio. The principals of dMark left, in part because they did not believe you can simply automate the relationship between buyers and sellers. There is a value to the interpersonal contact.

Quigo is demonstrating that the principle holds true for contextual advertising as well. While their transparency is important, I believe one of the greatest assets their platform has is the ability of the publisher to have direct sales and interaction with the advertisers. Instead of using the platform to create a vale between the two as Google does (and Yahoo!), they use it to help facilitate the interactions, with the added value of providing additional advertisers to the publisher if they choose.

Google’s push for innovation has been hampered by its dependence on the status quo. Now that they are big and mainstream with investors to please, they can’t do things that will put their current revenue at greater risk. Thus, change will only come when pushed on the market by other players. In this case, transparency and facilitating relationships between publishers and advertisers come at Quigo’s beckoning. Google follows in transparency.

The next acquiescence from Google will have to come in the form of control, allowing the publisher, through the Google platform to directly manage the advertiser base for its site. This is not what Google wants. They understand their importance in the advertiser relationship. But, if Quigo scales, pulls more major publishers, Google will be relegated to tertiary site traffic via adSense. I don’t think most people believe this is likely. But the dynamics of interpersonal relationships make this a very real possibility.

Despite 6 Sigma (and its predecessors) and all the business books’ protestations to the contrary, the world of business does not begin and end with efficiency. Yes, we must spend our dollars wisely. But, wise decisions start with trust, and trust is about relationships, not numbers. While Google understands this in its dealings with its employees, and with the work they do with us directly, they seam to discount it when looking at the interactions between other businesses. They can not assume their tools will be adopted over alternatives simply by claiming they are more efficient or effective. I guess the change here is that we now have alternatives.