Mike On Ads

A security firm, SecureWorks, has identified ads spreading trojans. Apparently the ads have specifically targeted job sites. Sadly they haven’t posted any details on the actual advertisement that is causing the trouble — I have sent them an email requesting some more details and hopefully they’ll respond and I’ll post the details of the ads here.

I find it kind of funny that the most interactive form of advertising has the worst reputation of all. Take this comment on my post about cookies:

your perspective on the ads seems very strange/scary to me. so a company gathers data about my surfing habits and whatever actions i do, and what do _I_ get in return? Longer load times, ugly flash clips, and bloody kilobytes of useless javascripts in every page.

Admittedly this is somewhat of an extreme reaction, but most consumers do have a very strong aversion to ads. And we as an indsutry are entirely to blame for this.

First and foremost we are often too lax with standards and create ads that can be highly annoying to end-users. Although premium publishers such as Yahoo or AOL are generally very careful about their ad quality many publishers and ad-networks allow highly distracting or even those “seizure inducing” ads that can be incredibly annoying. Yet even the best can be tricked, both AOL and Microsoft have been caught by the Errorsafe scam, which Sandi has documented here for MSN and here for AOL.

It’s not just “bad ads” that have caused this overly negative reaction to online advertising. This slashdot post complains about the latency of ads, a very valid concern when it comes to placing third-party content on a website. It is slightly ironic though that people will complain about waiting a couple seconds for ads to load when on tv we spent almost a quarter of our tv time being forced to watch ads. The thing is, even in a tech-savvy forum such as Slashdot, not a single person points out that all the services and content that they consume for free is paid for by advertising.

Of course consumers often may not even know who they have to blame for the ads that they are seeing. Take this rather typical request for help about “unwanted popups from yieldmanager.com”. For those of you that don’t know, yieldmanager.com is the serving domain that Right Media uses. This poor user thinks that his machine has been infected with spyware whereas what is far more likely is that the websites that this user visits intermittently show popup ads. Popups are of course closely associated with spyware/adware and tend to elicit a very strong negative reaction — it really is kind of ironic of course that popups often result in 10x the CPMs compared to traditional banner ads.

Last, but not least we have to talk about those terribly evil cookies. ClickZ covered this rather well in this article. Every “anti-spyware” program removes cookies and some flag them as “dangerous”. Of course there are privacy concerns, and perhaps adserving companies AND websites should be more open with what they track and how to “opt-out”, but cookies aren’t DANGEROUS, they just track what ads you’ve seen and what website you like. Here’s the little known fact — publishers are often the ones sharing the information and showing those popups, but the adserving companies end up getting all the flak. Instead of launching initiatives such as banning firefox publishers should keep an open dialog with their users about why they have advertisements and how they help to pay for all this free content.

If ads continue to have the reputation that they do we’ll just see a larger and larger increase in the # of users that use ad-blocking technologies. I think that Google is taking some good steps here with their recent blog announcement on their adserving tests. Unlike most companies, Google is very open and honest in their privacy policy about exactly what data they are tracking in their cookies and how they will use it. I suggest all privacy councils take a queue from them when writing their next privacy policy!

Education about what’s in a cookie is only the first step. Websites need to be more honest about the information that they pass along to third-parties and how that is used. Sites should also aim to educate their users that the reason they can consume content for free is because of ads.

The “Exchange” Buzz!!

Back in May I started a three part series on “The Ad Exchange Model” where I focused primarily on the technical benefits that exchanges bring to the online advertising industry. Since then exchanges have received quite a bit of press with all the recent acquisitions and the word exchange has reached buzz-word status, without much understanding of what it actually means. Perhaps most confusing is that many don’t seem to be able to differentiate between an Exchange and an Ad-Network. For example, compare these two quotes from iMedia and Ad.com, can you tell what the difference is?

“An ad exchange is a company that brokers online advertising by bringing publishers and advertisers together on a website where they can participate in auctions for ad space.” iMedia Connections — Ad Exchanges At a Glance

“Websites have ad space. Advertisers have ads. We’re the middle man – using our phenomenal technology to match ads to space.” Advertising.com Homepage

The ‘Nasdaq’ Analogy

One of the most common explanations I have heard goes something like this — “An ad-exchange functions just like the Nasdaq.” (this is in no way a jab at ContextWeb’s Adsdaq, this analogy is regularly used on all ad-exchanges). You like the Nasdaq right? Good, the Nasdaq is efficient has some fancy electronic trading and does lots of good things, which is why you should buy this ad-exchange. Oh, well of course! Ad exchanges bring efficiency just like the Nasdaq does, that makes absolute perfect sense! Try explaining an ad-exchange like this, it’s actually quite amusing. Generally what happens is the other person’s eyes glaze over slightly and he starts nodding as if everything has been made extremely clear even though he still has no clue what the ad-exchange actually does. You see, the Nasdaq analogy really doesn’t make sense but nobody wants to sound stupid and say “I don’t get it”, so they let it slide and remain confused.

So why doesn’t it make sense? Well, we could argue semantics of stock markets versus commodity and future exchanges — but who cares. The real problem is that the Nasdaq analogy works just as well for Advertising.com as it does for Right Media, adECN or AdsDaq. The Nasdaq is a mechanism which enables people to buy and sell stocks. Ad-networks are mechanisms by which people buy and sell ad-inventory. If this is surprising, it shouldn’t be — an ad-network is a mini-marketplace, very similar to an exchange or stock-market. The whole reason we have them to begin with is to enable thousands of sites to work with thousands of buyers. Can you imagine Netflix writing individual checks to the thousands of different sites where their ads are displayed?

So what’s the difference

Even though an ad-network may perform many of the same functions as an exchange there is a subtle difference. The network operates and controls all aspects of the mini-marketplace whereas the exchange is an “agnostic” platform that many buyers and sellers use to run their own businesses. In most ad-network limits each transaction (or ad-impression) to three parties: a buyer, the network and a seller. The exceptions to this would be networks such as Tacoda where a fourth data-provider might receive a cut of the transaction based on information about the user that the network provided. Of course in reality online advertising is far more complex than that and a single online ad impression can involve far more parties.

That’s where the exchange model shines — instead of one party “operating” the exchange on behalf of a number of buyers and sellers, the exchange provides a single technology platform upon which many companies — advertisers, publishers and networks — can buy and sell ads. Each impression can have anywhere from zero to many middlemen.

Who cares?

First off, the exchange is many companies versus one in the case of the ad-network. The platform is an ecosystem that supports a large variety of business models which results in more innovation and competition. Then there are certain basic technical benefits from having multiple participants in a transaction use the same platform, something that isn’t possible with an ad-network. For example, a proper exchange removes the operational barriers that have limited access to inventory in the past are eliminated in an ad-exchange. This enhanced the liquidity of the marketplace, which results in higher and more stable rates for publishers. I’ve covered most of these benefits in detail in my ad-exchange series.

Of course some of the above is still somewhat theoretical as we are just now starting to see the exchange model mature. As the model grows, most standalone networks will find life difficult without some level of integration with one or more of the coming exchanges.

In Parts II & III of this series I’ll talk about some tactical steps Publishers and Advertisers can take to leverage exchanges as either sources of inventory or pools of advertisers to maximize ROI and revenue and perhaps some bits on how Networks can different themselves in the coming landscape.

I was introduced to the online radio Pandora this afternoon. Pandora is the human interface into the “Music Genome Project”, a 5 year project that had human music experts classify every single piece of music created (except for latin & classical) since 1900 with various characteristics. Specifically: arrangement, beat, form, harmony, lyrics, melody, orchestration, rhythm, syncopation, tempo, vamping and voice (source). The Pandora way it works is remarkably simple. You start your own radio station by seeding it with a song. Pandora then analyzes all the genome aspects of that songs and finds songs with similar aspects to play. Even straight off the bat I found the selection to be absolutely amazing. Each song you hear you also have a chance of ‘training’ Pandora by giving it either a thumbs up or down. I threw 7 very varied and rather random songs at Pandora and was absolutely shocked by the quality of the selection.

Now here’s my question — to me music seems so much more complex than online ads. Lets think about this for a minute… lets say on average 50 billion ads are shown a day. That means that this year alone we’ve shown about 17 trillian ad-impressions. Lets be somewhat pessimistic and say an average click-rate of around .1%, or 17.8 billion clicks. How is it possible that we as an industry haven’t been capable of showing users more relevant ads?

One of the interesting things of course is that Pandora requires on a “seed” and then subsequent feedback from a user. Is the future world one where you start your day with a search on Google, which then gets treated as a ‘seed’ for your days ad-stream until you search for something different? Can we ever convince consumers that with only a little bit of their help we can greatly improve their online experience by showing them ads that align with their interests?

As an added note, I must say Pandora does a terrific job of integrating advertising with the site. Check out this “Bose Ad” that I received midway through my stream:

Most people I know fall rather squarely on either the “business side” (marketing, sales, account management) or the “tech side” (engineer, support, product management, etc.). If you’re in doubt, there’s a rather simple test that almost always works to identify who is who. Imagine this situation, an ad is shown on a website. The advertiser is “Cingular”, the publisher is the New York Times. There are two brokers involved in this transaction, ad-networks ValueClick and Advertising.com. Advertising.com has a direct relationship with the New York Times and daisy-chains the impression to ValueClick which works directly with Cingular. Each party is using their own tracking system.

I can draw this in two different ways:

The difference here is very subtle. Which one do you think is correct? #1 or #2? If you were at a white board explaining this situation to someone how would you draw it? Think about it for a second before reading on.

In my experience, those who pick #1 are tech-inclined whereas picking #2 indicates an inclination towards the business side of things. Why? Pretty simple actually. This ad-impression can be thought of in two ways, a series of financial transactions between four business entities OR a series of HTTP redirects. Technically the impression starts when the user visits the New York Times and flows towards the advertiser via a series of ad tags to Ad.com, then Valueclick and finally Cingular. Financially the money starts with Cingular and flows first to ValueClick, then Ad.com and finally to the New York Times.

A tech person tends to follow the flow of the ad-call whereas a business person tends to follow the money. When I first started in online advertising I would only draw #1 on a white board, but since have learned to change the way I draw things depending on who I’m talking to.

So take this scenario. Your company just signed a huge deal with Avenue-A. You’ve just gotten a million dollars to spend for the quarter, it’s the big break you’ve been waiting for! You send your ops guy Marvin to coordinate the details with Avenue-A’s people and open a bottle of bubbly to celebrate. A week later you pull some reports and realize that you haven’t run a single impression yet. You yell out —

   ”MAARVVIIIIIIIIIIIIN!!!!!”.
   ”Yes sir?”
   ”Why the hell isn’t the Avenue-A deal running?”
   ”Oh, we couldn’t get click-tracking to work yet. I’ve got a ticket into support.”

Sound familiar? How about trying to get behavioral pixels live on a publisher’s page? Enabling age & gender passing on ad-calls for custom optimization? These mundane seemingly simple tasks can take weeks. Why? normal people using complex technical products.

Read this quote from the MSN Games 300×250 ad standards:

(A) Click-through tracking is not available on the following advertising elements:
   • HTML advertising elements that use method=”POST” for form submittal.
   • Rich media elements that use embedded or compiled URL information (Macromedia Flash creatives that do not use the FS command, for example).
   • Third party served HTML (IFrame) campaigns.
(B) Cache-busting is available automatically for pre-approved third party served agencies, others by request only. Exception: Third party served click URLs for hard-coded placements (text links, etc.) are not cache busted. Therefore, MSN click data for hard-coded placements using third party served click URLs may not match click data from the third party agency. Please enter the Cache busting tags; MSN is not responsible for entering these tags.

I can’t be the only one that thinks that this is absolutely ridiculous. How can it possibly take five days to get an ad live? This is a process that should take mere hours! Why should people know the difference between method=”GET” and “POST”? Your average media-buyer probably has a degree in English, your ad-ops guy? History? Philosophy? It shouldn’t be the least bit surprising that these people have so much trouble doing what those of us with technical knowledge may find simple.

By nature what we do is technical, it is called online advertising! Purely from a technical perspective we cant buy or sell inventory without snippets of HTML that enable communication and highly complex adservers to do decisioning and optimization. What makes matters worse is that most ad-impressions involve two adservers, if not many more. The way things work today, these systems can’t talk to each-other directly and all communication must pass through the browser, which only reads HTML, XML, Javascript, Java, etc… more technical things!

Yet, just because adservers can’t communicate without HTML, Javascript and redirects doesn’t mean that users should have to worry about these details. Come on! Think of WordPress or Blogger, both services that have brought web-publishing to the masses. They have abstracted almost all technical aspects of blogging allowing anybody to post their thoughts online — why can’t we do the same in our products? I’m not sure I have a solution right this second, but perhaps we should begin by re-examining what is considered “normal” today. Here’s my list:

1. Knowledge of HTML is needed to enable clicktracking/li>
2. 5 days is considered normal and acceptable turnaround time for trafficking third party ads
3. Passing user data between systems involves knowledge of querystrings
4. Adservers have vastly different standards, heck, users need to know whether to encode or not
5. None of the above can be implemented without a “test” to confirm it really works

Most of the above stem from the fact that we have many different platforms, a situation that is unlikely to change in the foreseeable future. Sure, if both the advertiser and the publisher are using one system things are easy — DART has internal redirects, Right Media has linking and I’m sure others have or will invent similar concepts. The problem is, there will always be advertisers and publishers that will work with all three systems — which the way things look today means more redirects, click-tracking and long delays to activate campaigns or setup tags.

So product managers, architects and engineers, I think it’s time to put on your thinking caps to figure out how systems can better work together to make the day to day lives of the people who actually use said systems much easier. A million dollar Avenue-A deal should never hinge on a trafficker editing the HTML of the ad-tags to enable click-tracking.

Seems 50-cent wasn’t too happy about an ad on Myspace that asks the user to shoot the rapper to win a free something or other (xbox, ps3, whatever). Article can be found here.

I remember this ad very well as I used to manage some publisher relationships when this first came out. If I remember correctly there were about 5 different variations on the ad, each time a publisher would ban one iteration the advertisers would come out with a new slightly less offensive one — it started with a bloody ‘shoot the rapper’, then the blood was removed and then various “beat the rapper” and finally I think it settled on “guess the rapper” with a multiple choice between various famous rappers. Needless to say publishers weren’t too happy with any of the above.

What this shows is how much advertisers are willing to push the envelope. The more offensive an ad, the more likely someone is to click on it. I know, strange but true. And guess what, once you’re on the landing page it’ no longer about the rapper but about getting a free Xbox or PS3 by completing all sorts of offers. If you don’t care about your “brand” why care if you offend your users?

I wasn’t able to find this ad running anywhere so I’d much appreciate it if someone could send me a copy or link to the ad.

Just saw this ad on CNN.com. It takes the traditional “punch the monkey” ad to a new level — a game that’s actually semi-enjoyable and not overly obnoxious. They’ve even added a “tell a friend” button so that you can email the “game” to your friends. I’m guessing that the conversion rate on this ad is pretty high.

-Mike

I’ve been quite a few emails about my ErrorSafe page, both from publisher and end-users, but not a single one from an ad-network. Interestingly I’ve gotten two separate comlaints about ErrorSafe on CareerBuilder.com. The first complaint I received was from April 22nd:

Just came across this information after receiving an odd message from drivecleaner.com. I was on Careerbuilder.com at the time, which doesn’t surprise me since my resume was “picked up” by some email scammers (phishing for SSN#s I guess) from that site recently. I have also been browsing EBay quite a bit lately. What exactly do I have to fear, as a consumer, from these heavy-traffic sites? I rarely click on the types of ads you have shown here, but now I NEVER will. I’ll just go directly to any website I want to visit. I work on an Apple MacBook,which I believe to be far less susceptible to viruses, and this drivecleaner showed up while I was browsing in Firefox. Any thoughts?

Then this morning, I received a second one:

Hello, I know very little about techie things, but i am on a mac and get the errorsafe messages on the careerbuilder website.

So, I thought I’d take a look, and sure enough what do I see?

Apologies for the crappy windows paint outlining but what you see there is a “mysurvey4u.com” ad on careerbuilder.com. If this name doesn’t ring a bell, I’ve already identified this as a fake front for running Errorsafe/Drivecleaner/Winfixer ads on my Errorsafe Page. Of course I have a tracer open, and what do we see? (Displayed URLs edited for clarity)

14:03:40.937[344ms][total 344ms] Status: 200[OK] GET http://oascentral.careerbuilder.com/RealMedia/ads/adstream_sx.ads/careerbuilder.com/JobSeeker[...]

14:03:41.640[375ms][total 375ms] Status: 200[OK] GET http://a248.e.akamai.net/[...]/oascentral-s.realmedia.com/RealMedia/[...]/Mysurveys_Burnads_120_ROS_Jul07[...]

Now I”m mostly guessing here, but it looks to me like CareerBuilder is hosting this creative themselves using 24/7 RealMedia’s Open Adstream adserver. Looking at the URL I’m going to guess that the creative for MySurvey4U was supplied by “BurnAds” — which from their website appears to be yet another online advertising/marketing/ad-network.

If you happen to have a contact at CareerBuilder they could probably use a little education on ad-safety.

Users, not advertisers

A friend directed me to an interesting post on Publishing 2.0 — ‘Ad Platforms vs. Ad Networks: Who Controls The Advertiser Relationship?‘ — using the new Time Inc/Quigo relationship as an example the author argues that “As more advertising dollars pour online [...] whoever controls the advertiser relationship holds all the cards”.

This comment made me think for a while about the publishers role in online advertising. Is it really all about building advertiser relationships? Most publishers that I have talked to over the past few years split their media into two buckets — premium & remnant. One sales force focuses on agencies and another on ad-networks and direct marketers. Often the second salesforce is one or two people who spend their days prioritizing networks.

Although advertiser relationships matter a lot on an individual deal basis, but I don’t see them as something that can be “controlled”. Think about it — lets say you have a great “premium” sales guy, Joe, who has solid relationships with two large agencies. He used to work at one and has a whole set of friends at the other. Quarter after quarter Joe just keeps bringing in tons of business. Well, what happens when Joe is poached by another company and leaves? Was the strength of the relationship based on Joe or on the company? What if Joe calls up the agencies from his new job and tries to convince them to move their dollars to his new gig? Who will the agency go with? It’s not that relationships aren’t valuable, but they are largely based on personal relationships between people — something that is far more fragile than a deep technical integration.

If it’s not advertiser relationships, what should publishers focus on? First and foremost every publisher should focus on increasing the user-experience on his properties, attracing more and higher quality users and collecting more information about said users. I think Niki Scevak has it right in this post — Running an Ad Network — where Niki argues “You’ve heard me crap on about the business model of ad networks enough but here it is again: over time all the value goes to the person who owns the consumer relationship.“. It’ kind of like that movie Field of Dreams, “If you build it, he will come”. Not that you don’t still have to work to bring in the advertiser dollars, but quality users make selling a lot easier.

The means of trading

So lets look at the example Time/Quigo cited above. Pre-Quigo, Time used Adsense because it enabled both contextual based buying and smaller credit-card based advertising on Time Inc. Adsense was able to provide a certain level of efficient buying and hence was able to pay higher rates than other advertisers. Citing from the Media Post article on the subject, Quigo states that “This is about publishers who want control over the ad-serving process, rather than outsourcing it to a blind network.” Well, I call bullshit!

This isn’t about Time gaining control — they’re simply transferring control from one party to another. Before the deal they had control over their users and inventory, and now they still have control over their users and inventory. The only real difference is that instead of sharing some margin with Google & Yahoo they are sharing some margin with Quigo. Quigo is getting exclusive control over the contextual advertising on Time Inc. Not convinced? Why don’t we take a look at the terms of service of a contextual buy on Time Inc

1. Display of Advertisement. Advertiser agrees that Quigo may display the advertisement Advertiser places in the AdSonar interface (the “Advertisement”) on the Quigo Network through the AdSonar Service and its affiliated sites on which Quigo places AdSonar advertisements

That’s right — this isn’t an exclusive Time Inc contextual advertising account I just signed up for, it’s a full-fledged Quigo account, it even works when I go login through the regular Adsonar login portal. So Time transferred control over the means of serving the ad from Google & Yahoo over to Quigo. So why do we care? The thing that is up for control is the means by which ad-impressions are traded or transacted. In the example above, Quigo may have tricked Time into thinking they were getting more control but really it’s three giants battling over one method (contextual matching) of serving ads on Time Inc’s inventory.

I find it hard to believe that this is financially the best move for Time. If Google or Yahoo pays a higher rate than Quigo, do you really care whether you are “selling directly” or not? Which brings me to my last thought of the day. Assuming a publisher has built up a quality user base, the focus should be on empowering all advertisers to best monetize the available inventory. Who cares if it’s Yahoo, Google, Quigo or even Bidclix! Whomever pays the most should get the inventory. If today that’s Google, great, but if tomorrow it’s Quigo then give more to them.

Next up

I realize that this post leaves a lot of questions unanswered — in my next posts I’ll talk more about maximizing revenue by pricing networks efficiently and also about how ad-networks, exchanges and technology providers are fighting for control over the means by which ad-inventory is traded.