Lotame

At Lotame, we are constantly striving to stay ahead of the curve and innovate. Since our inception, we have aggressively sought out and developed new, valuable metrics to measure the success of individual online advertising campaigns, as we do not believe (and have never believed) that clicks and conversions tell the complete story.

After much exploration, research, and analysis over the past few years, we have also realized there is no “magic metric”. In other words, there is not one individual metric that has the capability of measuring success for all campaigns. A key reason for this is due to each advertiser or campaign having a different goal. Some advertisers just want users to become familiar with their brand; others want users to purchase their product; while others want users to watch their television show or movie. One metric cannot accurately accomplish measuring each of these goals at the same time.

A year ago, we created our own single, unified metric for every campaign. It was a combination of every metric we look at in measuring the success of the campaign. However, there was an issue with creating such a metric – how does an advertiser compare the performance of their campaign by media buy/site? If only one buy/site uses the metric, then you cannot compare it across all buys/sites. Since the metrics we use incorporate our technology, it is not always easy to reproduce for other buys.

As a result, we measure the performance for each campaign individually, which also allows us to customize what we measure and report on for each campaign/advertiser. At Lotame, we have always prided ourselves on being 100% customizable for advertisers. We create completely customizable targets – no buckets. Therefore, creating completely customizable performance measurements seemed like a natural progression. While there are typically only a few possible end goals of a campaign (awareness, conversions/purchases/intent to buy, intent to view, buzz generation), many advertisers have smaller, intermediate goals. These smaller goals are fairly unique to each campaign. Therefore, we measure each of these goals as well as analyze the other metrics we deem to be valuable for their end goal. We want to stress that the smaller goals only exist as a means to measure the larger, end goals. And that is where our knowledge and experience at measuring the end goals comes into play. For instance, an advertiser may run a campaign asking users to upload photos for a contest. The likely reason behind running such a campaign is to generate awareness of the product/brand. We will not only measure the performance of the contest entries, but also various other metrics (i.e. Lotame’s patented Time Spent metric, the amount of time a user spends on the micro-site and/or the number of pages consumed on the micro-site, the responses to one-question in-banner surveys, and various engagement metrics with the creative) to get a more complete picture of awareness. The advertiser should understand that their contest is only one small portion of measuring the awareness of their product/brand. Lotame will complete the story with our knowledge, technology, and customized performance metrics.

We will continue to strive to identify and evangelize the best metrics possible to enable our clients to set appropriate goals for their campaigns and gain optimal insight into campaign performance. In doing so, we may find our way back to some uber-metrics, entirely new metrics or most likely, combinations of metrics which can be customized to match the specific requirements of each campaign. We’re fortunate to have great partners in this quest – our forward-thinking brand, agency, and publishing partners and companies including Vizu and Dimestore. Ultimately, the continuous development of new and scalable metrics will provide our clients with the best approaches for guiding and measuring their online advertising campaigns.

Written by: Doug Pollack is a Senior Business Analyst with Lotame Solutions. Doug has been responsible for the development of leading, and innovative Key Performance Indicators as well as new metrics for digital media. Prior to Lotame, Doug worked for a division of the Army where he leveraged his statistics background to analyze the aging of tanks operating in Iraq and Afghanistan. Doug is available at doug [at] lotame [dot] com.

one-question in-banner surveys: Surveys have gotten a bad reputation recently – mainly due to poor response rates and inaccurate results. However, the surveys we run are not affected by these issues. The types of surveys that are affected are the long surveys that force users to stray away from the website they were on. Users, especially those on social community sites (like Facebook, MySpace, and Bebo), do not want to be taken away from the websites they are on. These long, in-depth surveys do just that. They take the user away from the sites for a very long time, as they typically have 10 to 20 pages of questions. Therefore, the response and completion rates of the surveys are quite poor, leading to limited responses. With limited responses, results are typically inaccurate. However, the quick, in-banner, one question surveys we run solve these problems. Users can quickly take the survey without leaving their community. This leads to higher response rates, more responses, and more accurate results. In fact, Lotame has just finished a study testing the accuracy of the results of the surveys. We ran a survey measuring intent to view, which asked users if they planned on watching the season premiere of a popular TV show. We then re-surveyed the users who responded with a 2nd survey asking if they actually watched the premiere. 83.3% of users who responded to the 1st survey saying they would “Definitely” watch the premiere, did tune-in and watch the premiere. 74% of users who responded to the first survey saying they would either “Probably” or “Definitely” watch the premiere, did tune-in to the premiere.

On Monday June 1st 2009, Lotame’s very own Scott Hoffman (Me) will speaking on a panel at the EconAffinity Conference in New York City. The event kicks off internet week, and is produced by ContentNext. For a 25% discount of the price of tickets use the code AFFINITY09 in the checkout process.

Last years conference was great, and we expect this year to be even better. I personally will be participating in a panel moderated by industry legend Wenda Harris Millard entitled “Online Advertising: Is Turning to Behavioral Targeting and Bigger Ads the Answer?”

Here is a description of the panel topic:
In the last few months, Google announced it would test behavioral targeting and the Online Publishers Association unveiled new ad formats that include larger sizes. Are publishers recovering much-needed revenue by incorporating interest-based advertising and super-sized ads into their online strategy? How are they working to retain users who find larger ads obtrusive and behavioral targeting invasive?

Confirmed panelists:
Scott Hoffman, CMO, Lotame
Pam Horan, President, Online Publishers Association
Michael Keriakos, Co-Founder and President, Waterfront Media
Dave Morgan, CEO and Director, Simulmedia

Related articles by Zemanta

• Google to Start Behavioral Ad Targeting (businessweek.com)

Data is King. If you don’t believe me, consider this:

• Rental car companies and insurers are refusing service to people with poor credit scores because data mining tells them that credit scores correlate with a higher likelihood of having an accident.
• Nowadays when a flight is canceled, airlines will skip over their frequent fliers and give the next open seat to the mine-identified customer whose continued business is most at risk. Instead of following a first-come, first-serve rule, companies will condition their behavior on literally dozens of consumer-specific factors.
• The “No Child Left Behind” Act, which requires schools to adopt teaching methods supported by rigorous data analysis, is causing teachers to spend up to 45 percent of class time training kids to pass standardized tests. Super Crunching is even shifting some teachers toward class lessons where every word is scripted and statistically vetted.

(An excerpt from Super Crunchers, by Ian Ayers. Great video of him below - bottom of post. Worth a watch.)

The data validates the fact that data is King.

In the world of Online Advertising, marketers, web publishers and technologists alike are beginning to realize that data is hugely valuable. In the past, advertisements were served to websites based on the content of the website. Today, advertisements are served to websites based on data that is informing the ad server to show that advertisement. As a result of this paradigm shift, the digital media community is beginning to realize that media and data are separate commodities. Where you see your advertisement is very different than how or why that advertisement is there in the first place.

Being able to leverage the power of this online data is great — but what does it really mean? Let’s take a look at the type of data that is available for digital marketers.

There are two types of data we should look at.

1. Singular Data Points
2. Multi-Variable Data Points

A “Singular Data Point” is a piece of data that is very cut and dry. It is a binary 1 or 0, “yes” or “no.” For example, if I recently went to an e-commerce travel site, configured a flight for 2 people from NYC to LA, and clicked “view price”, then you could make the argument that I am very interested in flying in the near future. The instant I click the “view price” button, a single data point can be obtained classifying me as “someone interested in domestic travel.” This may hold a ton of value for marketers at the bottom of the advertising funnel (see diagram below) who are looking to convert on a very specific type of user for a very specific product, and has been, still is, and will remain extremely valuable. Some examples of “Singular Data Points” include: Purchase Intent or In-Market Data, Age & Gender and Household Income.

These are data points that explicitly and intuitively describe certain consumer attributes. They are more or less declared characteristics or definitions of a consumer.

“Multi-Variable Data Points” however, tell a very different story. Imagine for example that someone, who happens to be female, 24 years old, and makes $40k/year, went to a social network, checked her inbox, wrote on a friends profile page about last night’s TV show, left that site to read and comment on an article about the New York Giants, uploaded a video about comedy, checked the price of a flight from NYC to LA, and finally, viewed another friend’s photo album who happened to have just returned from a trip to LA. What does this social data say about that user? Nothing? Is it completely arbitrary and meaningless? In fact, it’s exactly quite the opposite. Being able to leverage “dozens of consumer-specific factors” in real time gives a marketer the capability of executing and deploying various tactics such as:

• Influencer outreach strategies
• Creative optimization for advertising units
• Deeper visibility into brand engagement opportunities
• Custom optimization for various and wide ranging back end performance metrics
• Statistical data analysis and research
• Comprehensive search strategies
• Informative and cost effective media buys

Moreover, this social data presents huge opportunities for marketers looking to reach consumers at the top of the advertising funnel, but still has applications toward the bottom of the funnel as well.

Today, there is much debate around the value of social data vs. purchase intent data. How and when it should be used? How much should each cost? How should it be sold?

If we look at social data and purchase intent data as it might be applied in the advertising funnel, it would probably look something like this:

Now let’s consider social data and purchase intent data in terms of value vs. time. If we consider our previous example, we can make the argument that purchase intent data is hugely valuable for a short period of time. If I am looking to book a flight within 2 weeks, that does not necessarily mean I would like to travel a year from now. For a travel related company, that 2 week data point is again, hugely valuable.

Now if we consider our social data example, by demonstrating undeclared and implicit behavior over an extended period of time, we can make the argument that this data has tremendous value indefinitely. If we were to graph social data vs. purchase intent data on a value vs. time graph in today’s environment, the graph might look something like this.

But as companies become even more sophisticated, develop better applications of technologies and data mining, then someone’s day to day behavior might prove even more valuable in the future. This is certainly true when it comes to “creative optimization.” It is the idea that a creative ad unit showed to a consumer will be directly linked to their “social data” and that the consumer GENERATES OR INFLUENCES some form of content, message, conversation, or engagement. “Consumer Creative” – or “Social Creative” is the next leap of value to marketers, and as we evolve in the creative evolution of social media/data, the value vs. time graph will probably look something like this.

At the end of the day, having the ability to make educated decisions using comprehensive data sets is what will differentiate businesses from their competitors. It will give the forward looking organizations a way to stay relevant, efficient, and strong, while other companies continue to use outdated and inefficient methodologies.

And as these technologies evolve, become more sophisticated, and create incremental value for marketers, web publishers and consumers, the “data” companies will have a responsibility to uphold the privacy rights of all parties involved. These safety and privacy measures will become, and are already, inherent features of the technologies, because with the power of data comes the responsibility to collect and use the data in ways that provide appropriate protections for user privacy. Responsible industry members will continue to develop practices and policies that can work for marketers, publishers and consumers in this arena (we are one of those companies).

Nevertheless, if companies and marketers want to make the best decisions possible for their clients, they should consult the data for the answers and throw the guess work away. In our world of digital media, find your exact target audience and save money on wasted impressions or eyeballs. It’s all about the data, as well as harnessing the power of consumer driven information.

Co-Authors:

Andy Monfried - CEO/Founder of Lotame

Dan Reich - Business Development of Lotame

Good friend of Lotame, and a member of our Board of Directors, Satya Patel, just wrote a great post on his blog entitled “3 reasons that data will save online advertising”, after reading this I smiled big and long! His post is so good, well thought out, and smart that I thought that I would copy it word for word. It is too important of thinking to not share with you. I am not sure about the “blog etiquette” here, so please post all comments on this blog entry back at Satya’s blog “Venture Generated Content”

3 Reasons why data will save online advertising

It’s been nearly 15 years since Rick Boyce and HotWired famously popularized the use of banner advertising campaigns as a model for generating revenue online. Since then, there have been many, significant innovations in online advertising, including new ad formats, new pricing models, new targeting technologies and new metrics for effectiveness. Yet the value of online display advertising is being questioned now more than ever before, particularly in the current economic environment. Numerous organizations are projecting that online display advertising spend will be flat or slightly down in 2009. Growth is expected to recover in 2010, but at much lower rates than earlier in the decade and than search advertising. But the explosion of data and its increasingly effective use hold great promise for online display advertising. There are many types of data for online advertising, including keywords, contextual, behavioral, semantic, demographic, psychographic and social. The relative value of each of these forms of data is still an unknown, but I believe that the value (and cost) of data will soon exceed the value of inventory, which is already deteriorating.

Here are three reasons that the use of data will save online ads and help restore their growth.

- Data makes media buying easier: Data from comScore, the IAB and others suggests that while the top 50 online publishers only account for 25%-35% of user attention, as measured by page views or time spent, they represent about 90% of online advertising spend. Why is that? As I’ve written before, the job of an online media buyer is seemingly impossible. Audience fragmentation, the proliferation of ad networks and the emergence of ad exchanges have created incredible amounts of complexity in the marketplace. Learning about all of these sources of inventory, let alone buying from them, is an unenviable task. On the other hand, buying from large, known publishers is simple. This is the default behavior for many online media buyers because it doesn’t entail extra effort or risk. Further, the buying of traditional media, rightly or wrongly, is done largely based on gross rating points, viewership, circulation, listenership, etc. Media buyers purchase audiences at scale. In the online world, media fragmentation has made it a necessity to buy from multiple places to achieve desired scale. Data allows traditional buying behavior (again, independent of whether it’s good or bad) to be replicated online. Data enables media buyers to purchase a specific, consistent audience at scale across many different publishers. Data makes the jobs of media buyers easier, allowing more dollars to be spent online.

- Data increases the value of remnant inventory: Somewhere between 30%-40% of online ad inventory at most major publishers goes unsold by their direct sales organizations. That number is closer to 80-90% for most social media sites, the fastest growing segment of inventory and the one with the most ad effectiveness challenges. Remnant inventory is the direct result of highly ineffective ads that are not relevant to the consumer. There was a time when NYTimes.com could sell its inventory because of the association with its brand. That time is long gone as metrics have told advertisers that they are not earning a return on their dollars. Getting value from advertising on social media, where consumers are largely not engaged in commercial activity, is even more difficult. And inventory, both premium and remnant is increasingly being commoditized by the ad exchanges. Effective use of data for targeting (with more engaging creatives) the right audience yields better ad performance and generates real value from remnant inventory. In the end, today’s gap between demand and supply diminishes as data-defined audiences, rather than impressions, are being purchased.

- Data is available to all: The traditional ad agency model is widely recognized as broken. The economics of the agency business dictate that they find more efficient and effective ways to engage consumers on behalf their advertising clients. Along these lines, agencies have come to realize that one of their greatest assets is their consumer and ad performance data. Data, in combination with more innovative creative, can target the right audience at the right time with the right conversation, interactivity and engagement. Publishers also see that it’s becoming more difficult to aggregate sizable audiences and to sell their ad real estate. Differentiation in the face of commoditization comes from their data. And ad networks know that they are in danger of being disintermediated unless they bring unique value to the both advertisers and publishers in the form of greater access to data or better targeting through data. Fortunately, all of these players have their own data assets and increasingly have access to data from traditional offline data vendors, such as Acxiom and TARGUSinfo, as well as from emerging online data exchanges, such as BlueKai (where I am an investor) and eXelate. The competitive dynamics in the online ad industry dictate that the various players leverage data to provide greater value to their constituents.

While data doesn’t solve all of the problems in the online advertising market, it’s clear that data is going to have a huge impact on the future of the industry. The companies that develop the platforms, tools and services to make it easier to aggregate, analyze and utilize data will be the next category of winners in the online ad market. More importantly, they will help grow the online advertising market for all of us. Even as the value of inventory decreases, the increasing use and value of data and the resulting greater sell-through of inventory will yield a larger online advertising market.

It is no secret that the ad economy is currently weak. Even the rosiest of estimates have paired back general ad spending, from anywhere to up a small % year over year to down. There is no question that as we head into the 2nd Quarter of 2009, CPM rates will continue to come under pressure.

Premium content sites have begun to surface some evidence on their strategy on how to weather the storm, by introducing “paid for” services. In the last 24 hours, The Wall Street Journal announced their Niche content subscription plans, and the Washington Post announced their “paid for” image service. There is precedent for this corporate behavior, this very same cycle (ad market collapse / increase in paid services) happened in back in the 1999-2000.

But what are sites doing that don’t have a premium offering that is compelling enough to bundle into a “paid for” service?

These are the advertising revenue pure play sites, which represent the vast majority of web publishers today. This is especially true within the social media/community/blogger category.

It is widely believed that these pure play sites are coping by simply adding more advertising placements and ad inventory. Short term, the net effect is that if you add an ad placement, you can generate extra revenue against that page/site. The formula is quite simple, let’s say a site had two placements before the downturn which dropped their CPM’s by 33%, without any traffic growth by adding a 3rd ad placement the site could generate 33% more revenue (than just allowing for 2 ad placements) which would equalize the raw ad revenue generation that the site delivers. Seems simple right? Not exactly.

If every site simply added an additional ad placement, the pool of available inventory would grow and the laws of supply and demand would take hold and CPM pressure would drive rates down, hypothetically by the same % that supply increased. The ad revenue flowing into the online ad economy is a constant in the equation.

So what is the answer? TIME

The answer is for publishers is to look at the value of the ad placements in terms of time that the ad was exposed to a consumer. We released some great data about the average exposure time of three of the most popular ad placements; the rectangle, the skyscraper, and the leaderboard. The results of the study showed that the rectangle sized ad had a clear advantage staying visible on a consumers screen for 13 seconds, 2.5 times more than leaderboards (5.4 seconds), and 6.8 times more than skyscrappers (1.9 Seconds).

From a marketers perspective, these are big findings. If the basic premise of online marketing is to drive brand imagery, create consumer demand, and propel purchase intent, then the ad that stays visible for longer in front of a consumer is more valuable.

If  a publisher were to plan ad placements based on the data that we released, then theoretically a publisher might actually be able to reduce the number of ads served on a page, and get an increased effectiveness of the ones that remained. To illustrate, if a web site currently has an equal mix of rectangle, skyscrapper, and leaderboard ads on every page, and they move to two rectangle ads, the time that those two ads would be visible to a consumer would be 26 seconds, up 28% from 20.3 total exposure seconds with the current three ad placements per page.

By looking at the time spent metrics, a publisher can drive value for marketers (and marketers will reward publishers with higher CPMs), remove clutter from current web page designs, and drive greater and more efficient revenue. It is a win, win, win, scenario.

If you are a publisher that would like to understand more about what Lotame does, and how we can help you grow your revenue in a smart way email me at scott@lotame.com