The Performance Marketing & Display Network Problem (Part 1)
As the former owner of a Comscore rated display ad network, and now the manager of a major performance marketing network, I seem to have unique insight into how networks and performance/affiliate companies work. Before owning the major network, I was actually a multi-million dollar affiliate of my current company. I’ve a great deal of experience in making CPA-based campaigns work on “brand CPM” inventory – and the real issues in dealing with running CPA/Performance marketing campaigns on CPM display inventory. Thus, as part of my tireless mission to inform and educate the public with my brilliant analysis, here is the first of a two part series on the issues of CPA campaigns and display marketing.
Yes, there are big issues with CPA campaigns run on display networks and display marketing in general. It should be pretty damn simple to be honest – a display network decides that they want to run CPA campaigns, they run them, they get paid per lead, per sale, per action. Some networks see the CPA campaigns as fillers, other have recognized that many of them convert higher than even their “premium” branded campaigns and depend on them for a significant amount of the inventory. Here at Affilaite.com many major networks work with us in order to not only fill their inventory, but for cash flow since our payment schedule is much better than most “agencies.” However, what should be a simple campaign – a CPA campaign, conversion based campaign that backs into a CPM campaign often ends up in a problem.
That problem is surrounded by the unique problem of the definition of “lead” or sale. In the 
performance marketing community, much of the marketing is purely lead or sale generated, and the affiliates and the partners that work are so tied into the lead part of the industry that they examine every lead sometimes. They understand what a “good lead” is on a campaign, for example and work with the network and the advertiser to generate those “good leads.” However, on the display network side, the networks are completely and totally into the eCPM generated. This itself causes a significant issue often when it comes to the networks getting paid, and the actual money generated.
It’s quite simple actually, for example, in the performance marketing dating industry. I’m going to use this example, specifically, because this happened to me a few years ago and is quite relevant to the discussion at hand. My network ran a few hundred thousand dollars of CPA advertising for a major dating company, through a major affiliate network. We ran banners for about 90 days, it backed out into 30 cent CPM and was a great filler for the network – and we paid quite a bit to our publishers on these banners. Yet, around day 90, the CPA network came back and told us that the advertiser was not happy with the “leads” and felt they were “fraudulent” or “bad” and was not going to pay the bill in full. The affiliate network decided at that point to only pay a part of the final bill, owing us about 1/4th of the total run.
I understood what a bad lead was, but at the same time, I knew it wasn’t my network’s problem. We just ran banners on our network which backed into a CPM on the network. Our publishers were not paid on a CPA, and the poor performance of the signups was purely circumstance. However, since the dating company was so integrated into the performance marketing realm, they saw the poor performance at the end of the run, when the campaign was growing as a sign of fraud, and the affiliate network in turn didn’t want to pay for those leads.
Here is where the two worlds completely clashed – we had backed out the leads into the CPM for our publishers, there was no way for us to figure out which publisher generated bad leads, nor even a way to remove them. If our publishers suddenly were told they had a lower paycheck because of non-payment of the leads, they would have come back with “Uh? You pay us on a CPM to run your ads, why should we be concerned about the advertiser’s performance.” Similarly, as a network, all we cared was about was how it backed into a CPM. We weren’t pushing this as a CPA campaign, just running the banners on a CPA. I felt inappropriately spanked for something that our network didn’t do wrong…
Having talked to several friends at major networks that basically refuse to take CPA campaigns from most performance based networks; they raise this issue over and over again. They want to run the CPA campaigns, but will not sign any contract that doesn’t pay for “bad” leads. They aren’t in the lead generation business, they are purely running banners. They see their compromise of running CPA instead of CPM, to address that issue of “bad traffic” and frankly don’t want to have to deal with what a lead means.
So my question to the audience is how we can solve this issue? Obviously with the economy as it is, more and more brand advertisers will be turning to CPA campaigns. The networks are more open to this, and willing to speak to more and more performance marketing companies about working with them. However we need to come to a common ground first in order to grow this part of the market.
Love your feedback.
?
13 Responses to “The Performance Marketing & Display Network Problem (Part 1)”
Leave a Reply
(C) 2010 Pace Lattin. It is the policy of this blogger to not edit or remove any content and comments, unless it is specifically attacking a protected group or irrelevant to the conversation, such as a spam. These are the OPINIONS of the respective writers, posters, commentators. All DCMA Notices shoudl be sent to pace@pacelattin.com
Any media on a CPM rarely ever backs into the target CPA…better to focus on putting together CPA partnerships and growing that aspect in my opinion. That way you never have to worry if you are losing money!
You need better back-end tracking. If you (either the Affiliate/Network/Advertiser) cannot track leads to the site or placement, than they shouldn’t have the right to only pay for a portion of the leads generated. Certainly all CPA advertiser agreements allow the advertiser to withhold payment for Fraud, but that term must be clearly defined and defensible.
Most network today are able to track leads not only to a publisher level, but a sub-publisher and in some instances the specific creative or placement. If you had better transparency into that, you could cut off your poor performing partners.
In your example, since you backed the CPA campaign into a CPM campaign for your affiliates, you are responsible for the performance to the client and for the CPMs to the network. I assume you made money on the conversion of the CPA to CPM, so that’s the risk you took. If, however you offered the .30 CPMs to the advertiser, they might have jumped on it and then the quality of the leads is their problem.
You simply can’t get around providing ROI for advertisers. That’s the bottom line. If a campaign does not provide ROI — whether CPA, CPM or CPC — the advertiser is going to pull it and go elsewhere. That’s something that publishers and even networks don’t always understand. They think that traffic alone is valuable and it’s not…it’s quality traffic that’s valuable.
Cum werk 4 eliteaffz.cum we hav hiest payouts
This is very similar to the argument I hear from publishers/networks that send fraudulent traffic and expect to not be held accountable for it.
"I sent you traffic! Our contract states X,Y,Z! It doesn’t matter that it’s bad, I want paid!"
I lulz at them.
This isn’t to say that there are not advertisers that close on "rejected leads".
LolZ: you aren’t addressing what display advertising is. It’s putting a banner up, period. THere is no "lead generating." The banners go on sites, people click, then they submit. Quality shouldn’t be the concern of the network at that point.
The problem is the terms of your agreements don’t match up. If your agreement with the CPA network matched up with yours, you wouldn’t have a problem. 90 day lag on getting paid and scrub is a terrible deal….shame on you for doing this to begin with. If you want to make these deals work, ALL terms on the Adv and Pub side have to match up. Last I check with Affiliate.com is you don’t allow scrubs from Advertisers……that’s why we can’t do a deal!
I’ve rarely seem any CPA/L campaign back out to a decent CPM. In the UK the decent CPM networks are trying to get into running more CPL as CPM is going unsold and CPL is better than CPA or nothing. Realtime validation is essential so they cannot get away with wanting to be paid on gross. Anyone serious about lead generation knows that only valid leads count. Networks are either ignorant or naive about having to validate leads so I use third party technology so only valid lead counts are passed back to the network in the response code. The result is that networks only do 15% of the leads they thought they could but the quality is high. However they are still obsessed by eCPM and the ad server rules the roost so most CPL campaigns on decent newtorks are never given enough inventory. I’m working hard to get all the potential publisher and advertisers in on lead generation talking the same language – that would be start!
I’m not a huge fan of the CPA model, as marketers are essentially vying for free advertising. If I’m reading The Economist, I don’t clip out a BMW ad and send it to their corporate HQ so that they can pay The Economist for the lead. It’s ridiculous.
On the practical side though–speaking from the network side–it’s something we’ll have to live with to supplement premium CPM advertising.
As for tracking good versus bad leads? I agree with Kyle in that "..most network today are able to track leads not only to a publisher level, but a sub-publisher and in some instances the specific creative or placement."
Tracking is key. That being said, all advertising is bound to result in a percentage of bad leads, which is not to say it is fraudulent. Fraud is another issue that the industry absolutely needs to protect itself from.
This is an issue I saw all the time while working as a RMX Network Manager at a company which was primarily a CPA Network. As advertisers would come in the contracts would start out simple. Then ever time a "CPA publisher" would jump on prices would drop and leads would be rejected. I really see it as the fault of the shady co-reg/incentive based sites and networks. They have put a sour taste in advertisers mouths. It seem now there is almost an industry standard scrub rate leads.
As a site owner I want everything in CPM. It’s easy to forecast and simple to calculate. CPC’s work to because I can easily track the clicks in real time and get an eCPM. Leads are harder I have to track them on a daily basis at best and then there’s charge backs, cancellations, and "bad leads" so really it takes weeks before I really know what I’m getting paid. Then as a media buyer working for a OTA I only want to pay for Sales or buy on a rev share.
I Now believe that CPA is the only way for both sides this wasn’t always the case but after running some CPA offers with eCPMs over $500+ I have to say it works for both sides. It requires better tracking and Transparency. Then ultimately the weeding out of Incentive driven leads and shaddy publishers. CPA’s are here to stay and they will work for everyone when properly executed.
So $0.30 x 50% = $0.15 cpm to your publisher. You have to love the CPA to cpm arbitrage game. So $0.15 cpm for US inventory and you get gaming, anime, social net, forum, and joke sites. The creative were salacious I’m guessing and generated high ctr. But nobody bought anything and in the end roi was poor for the advertiser. Let me tell you a secret – that’s because your inventory was crap and the extra exchange inventory you bought to increase the campaign size was even worse but nobody cared while the money flowed.
The alternative tell the advertiser the rate is too low and they cannot buy decent inventory at this price and have another adnet take their money.
Its pretty simple actually – somebody needs to be arbitraging models and taking the risk. If a publisher wants to be paid on a CPM and an advertiser only wants to pay on a CPA, somebody operating in the middle with good access to data needs to guarantee the CPM upfront and take the CPA and make a return on the spread. If those brokerage returns get too large one side or the other will be willing to switch models. If they don’t exist then we’ll always have 2 models because CPA is inherently impossible due to incomplete data for all types of advertising. I’d bet on the latter, since I’m pretty sure the CPA/CPM arb business is already very efficient. Some sites are better geared toward delivering brand advertising then they are at delivering CTR, just like Vogue is different than Autotrader.
Great article to quote in my college esee!Thanks