Today, we released our ninth MarketPulse newsletter, a regular publication utilizing DataXu’s proprietary data to uncover and reveal interesting trends in digital advertising. This month’s report analyzes campaign data from two key verticals to see whether DSPs can leverage programmatic buying to find more valuable customers.
The full report can be found here: DataXu MarketPulse – July 2011
Demand side platforms are giving marketers the power to radically improve their customer acquisition strategies. These solutions have brought dramatic efficiencies and increased return on ad spend through programmatic buying on ad impressions that are most likely to convert, based on an advertiser’s specific metrics. However, many advertisers rely on cost per acquisition (CPA) as a preferred metric of success. But not all conversions are created equal, and the value of a customer can vary greatly.
Sophisticated marketers are looking for ways to find not just new, but more valuable customers. Valuable customers are the ones who spend more or buy more profitable products. So we asked: “Can DSPs leverage programmatic buying to find more valuable customers?”
Next-generation demand side platforms, like DataXu’s DX2, leverage advertiser data to create a custom algorithm that not only decides which ad placement is most likely to convert, but also which ad placement is most likely to drive the highest revenue. To illustrate how this works in practice, the DataXu Advanced Analytics Group reviewed the spend data and value of customer purchases from recent campaigns for companies in the financial services and technology industries to measure the impact of shopping cart value optimization – which considers the value of a customer’s purchase, in real-time.
The analysis (below) shows that campaigns optimized for shopping cart value spiked compared to DataXu’s standard cost per acquisition (CPA) optimized campaigns; delivering a 30% better return on ad spend for the financial services vertical, and an 80% better return for the technology vertical.
The DX2 platform uses dynamic decisioning to increase budget toward converting customers likely to spend more money, and reduce budget on those likely to spend less. The analysis (below) also shows that campaigns optimized on shopping cart value converted 25%-45% more high value customers than standard CPA optimized campaigns.
“The Pareto Principle holds that 80% of a company’s revenue typically comes from 20% of their customers,” stated Sandro Catanzaro, VP of Analytics and Innovation at DataXu. “Yet many advertisers make the mistake of equally chasing all of them. Advertisers focused solely on CPA are likely wasting a lot of money converting lower value customers. Today’s sophisticated marketers are looking beyond simple conversions and reallocating much of their spend to convert higher spending and more valuable customers.”
DataXu’s MarketPulse explores the data that defines today’s digital advertising marketplace. Sign up here to have future MarketPulse newsletters delivered directly to your inbox.
Share on Facebook- Posted by Denise on July 21st, 2011
Today, we released our eighth MarketPulse newsletter, a regular publication utilizing DataXu’s proprietary data to uncover and reveal interesting trends in digital advertising. This month’s report revisits our May 2010 study to see how market maturation is affecting media prices in 2011. Read more
Share on Facebook- Posted by Denise on March 19th, 2011
Today, we released our seventh MarketPulse newsletter, a regular publication utilizing DataXu’s proprietary data to uncover and reveal interesting trends in digital advertising. This month’s report examines the industry skepticism surrounding owners of ad exchanges and whether they can operate as neutral brokers given their ownership of most of the media offered for sale on the platforms. Is there a discernable difference between these platforms and truly independent and neutral Demand Side Platforms (DSPs) who utilize automated technologies to span across many exchanges, valuing and buying only the highest performing impressions at a value price to the advertiser? Read more
Share on Facebook- Posted by Thomas Joyce on December 20th, 2010
Today, we released our sixth MarketPulse newsletter, a regular publication utilizing DataXu’s proprietary data to uncover and reveal interesting trends in digital advertising. This month’s report examines Brand Genomes and the performance impact of custom campaign models versus one-size-fits-all content and audience channels.
The full pdf is available here: DataXu MarketPulse_Brand Genomes_November 2010.
DataXu MarketPulse: November 2010
Brands make enormous investments to develop differentiation that drives success in the marketplace. However, when buying digital advertising, that differentiation is often lost through a generic media plan. One-size-fits-all content and audience channels that target “soccer moms” or “auto intenders” fail to address the unique traits of a brand and its consumers. The loss of uniqueness is compounded when ad networks reuse campaign data across competitors.
In contrast, the world of real time bidding has evolved beyond predetermined channels, producing a new set of insights that make it possible to create an advertiser’s own unique “Brand Genome.” Using impression-level decisioning across multiple targeting parameters, an advertiser can for the first time build a data model that defines exactly how consumers behave in relation to a specific campaign. Deciding who, what, where, and when – ads are placed based on where they are proven to drive performance; enabling advertisers to discover new prospects they didn’t know existed.
To demonstrate just how unique each Brand Genome is, DataXu’s Advanced Analytics Group applied different campaign strategies to three leading automotive brands and compared the conversion success rates.
From left to right, the columns below show the performance of running Brand A’s campaign using a run of network buy, an audience buy, a buy modeled on competitive B and C genomes, and finally, a buy based on Brand A’s own genome. The advertiser’s own brand genome campaign drove conversion rates 2-3X higher than traditional campaign strategies. Interestingly, using competitive genomes did little to increase conversions, further validating the need to develop unique media strategies to drive optimal campaign performance.
For the full MarketPulse report, you can download the pdf here.
DataXu’s MarketPulse explores the data that defines today’s digital advertising marketplace. Sign up here to have future MarketPulse newsletters delivered directly to your inbox.
Share on Facebook- Posted by Julie Ginches on November 17th, 2010
Today, we released our fifth MarketPulse newsletter, a regular publication utilizing DataXu’s proprietary data to uncover and reveal interesting trends in digital advertising. This month’s issue examines targeting and its impact on the scale and performance of a campaign. The full pdf is available here and below is a snapshot of the research.
DataXu MarketPulse: October 2010
DataXu’s Advanced Analytics Group studied 14 campaigns across industry verticals including Auto, Software, Finance, and Education. These campaigns were based on agency media plans that featured varying levels of targeting, including by site, context, demographics, location, and consumer behavior.
The data¹ indicates that campaigns with more targeting in the media plan not only produce a lower volume of conversions, but also worse CPA performance. Indeed, fewer targeting parameters (1-2) drove 3X better scale and 2.7X better performance than campaigns with more parameters (4+).
Key take-aways:
For campaigns with few targeting parameters on the media plan, action volume was 47 percent above goal, and CPAs were 55 percent above goal.
For campaigns with many targeting parameters, action volume was 50 percent below goal, and CPAs were 40 percent below goal.
For the full MarketPulse report, you can download the pdf here: DataXu MarketPulse_Targeting and Scale_October 2010 DataXu’s MarketPulse explores the data that defines today’s digital advertising marketplace. Sign up here to have future MarketPulse newsletters delivered directly to your inbox.
Share on Facebook- Posted by Julie Ginches on October 14th, 2010
Today, we released our fourth MarketPulse newsletter, a regular publication utilizing DataXu’s proprietary data to uncover and reveal interesting trends in digital advertising. This month’s issue compares exchange targeting strategies – Audience Targeting, Retargeting and Optimization – to find the real cost of the media.
The full pdf is available here: DataXu MarketPulse_Exchange Trading Strategies_September 2010 and below is snapshot of the report.
DataXu MarketPulse: September 2010
“Comparing Exchange Targeting Strategies: Following the Crowd Will Cost You”
This analysis was prepared averaging cost of media across campaigns for consumer electronics, financial services, software, and automotive.
Data Highlights:
- The cost of media for Optimized ROE is 1.2X more expensive than media for general ROE buys
- The cost of media for Audience Targeting is 3X more expensive than ROE (but effectiveness as measured by CPA was 2.7x better for Optimized ROE than for Audience Targeting)
- The cost of media for Retargeting is on average 3X more expensive than ROE
“The data shows what the smart media buyers know: Audience Targeting and Retargeting are effective, but the scarcity of placement opportunities causes media prices to spike and impedes campaign scale,” said Sandro Catanzaro, VP of Analytics and Innovation. “Optimization, however, is not similarly constrained as it works across a much wider swath of impressions, which enables more buying efficiency and scale.” For the full MarketPulse report, you can download the pdf here: DataXu MarketPulse_Exchange Trading Strategies_September 2010
DataXu’s MarketPulse explores the data that defines today’s digital advertising marketplace. Sign up here to have future MarketPulse newsletters delivered directly to your inbox.
Share on Facebook- Posted by DataXu on September 9th, 2010
Today, we released our third MarketPulse newsletter, a regular publication utilizing DataXu’s proprietary data to uncover and reveal interesting trends in digital advertising. This month’s issue focuses on a highly contentious issue in the online advertising space – Last Click Attribution.
The full pdf is available here: DataXu MarketPulse_Last Click Attribution_July 2010 and below is snapshot of the report.
DataXu MarketPulse: July 2010
“Last Click Attribution: A Simple Way to Misallocate Your Budget”
Taking a look at data for six client campaigns, the DataXu team identified some interesting attribution trends.
Data highlights
- For all campaigns, last click attribution ignored 97% of spend driving conversions — which often results in over-spending in search and re-targeting, and under-spending in display that drives demand creation.
- The recommended attribution period for short consideration products, such as CPG, is two weeks. In the campaign shown above, this window includes 90% of impressions that converted.
- The recommended attribution period for long consideration products, such as Insurance and Autos is five weeks.
- The length of time it takes to attribute 90% of conversions varies by 250%; each product requires its own distinct attribution model
Ultimately, no two campaigns are alike, and what drives their success is tough to predict at the outset, but incredibly valuable to understand.
For the full MarketPulse report, you can download the pdf here: DataXu MarketPulse_Last Click Attribution_July 2010
DataXu’s MarketPulse explores the data that defines today’s digital advertising marketplace. Sign up here to have future MarketPulse newsletters delivered directly to your inbox.
Share on Facebook- Posted by DataXu on July 28th, 2010
Today we published the second issue of the DataXu MarketPulse, a research newsletter designed to highlight trends in digital advertising based on DataXu’s own proprietary data.
In this month’s issue, we explored campaign performance and the impression attributes that are most highly correlated with conversions and the results were surprising!
DataXu MarketPulse, June 2010
Beyond Audience: What Drives Campaign Performance?
Since the dawn of the Internet, advertisers have been on a quest for the holy grail of digital advertising: delivering the right ad to the right consumer at the right time. This quest has yielded great insights about how to improve each of these ad dimensions. Consumer targeting is the current priority for some high-profile ad agency holding groups. But which factor is the most important driver of a campaign’s success?
With real time bidding (RTB), advertisers can get closer to the answer than ever before. Impression level data essentially maps the path to the grail. At MarketPulse, we like to look past the hype and let the data do the talking.
With 100s of millions of impressions served daily, we recently evaluated whether consumer, context, or creative attributes of ad impressions were most predictive of conversions.
The result? Each category was well represented, but creative attributes were the winner — correlating most highly with conversions for 48% of the campaigns.
“The data shows that a single campaign performance driver cannot be predicted with confidence in advance. This suggests that a more effective ‘media plan’ is one that instead responds to the nuanced, unique data each campaign generates. The implication is that brands and their agencies need systems that can glean the data and automatically adapt, rather than relying on a rigid, one-size-fits-all approach,” noted Mike Baker, CEO of DataXu.
Our analysis included nineteen large-scale, online display campaigns that ran for at least four weeks, across the leading RTB ad exchanges. In each campaign, we evaluated multiple impression attributes in the categories of consumer (who saw the ad), context (where the ad appeared), and creative (what the ad looked like) to see which attributes appeared most consistently across ad impressions that preceded conversions.
For additional results and takeaways from this month’s MarketPulse, you can download the pdf: DataXu MarketPulse_Campaign Performance Drivers_June 2010
DataXu’s MarketPulse explores the data that defines today’s digital advertising marketplace. Sign up here to have future MarketPulse newsletters delivered directly to your inbox.
Share on Facebook
- Posted by DataXu on June 28th, 2010
Yesterday, we distributed what we hope will be the first of many “DataXu MarketPulse” newsletters. Designed to highlight compelling trends in the digital advertising space, the DataXu MarketPulse was created in response to our customers’ requests for visibility and insights into the rapidly evolving world of biddable digital media.
DataXu MarketPulse, May 2010
Extreme Price Volatility Across Ad Exchanges Presents Opportunities for Buyers
DataXu analyzed the price paid for ad impressions across ad exchanges over the past 30 days, and discovered that the average daily price varied by over 100% during the period.
This analysis is drawn from our leading cross-exchange advertising optimization platform, which evaluated over 175M unique user visits and hundreds of billions of ad impressions from all of the major suppliers of biddable display advertising during the measurement period.
“With the tremendous growth of exchanges for bidding on digital display ads, we are seeing a persistent price volatility across all points of supply,” said Mike Baker, CEO of DataXu. “This presents innovative advertisers with a huge opportunity to boost media effectiveness and efficiency. But to do so, they have to move beyond static media buying to dynamic decision systems that keep pace with the volatility.”
Advertisers know that exchange-traded inventory varies in value, but this significant volatility in prices for fungible placement opportunities may serve as a wake-up call. When it comes to display ads, are you confident that you are getting what you pay for?
Over time, across large campaigns, such price variances can have a significant impact on ROI, if an advertiser is not employing the right analysis and tools to ensure that their higher cost impressions are also the highest value impressions for their specific goals.
The Average Daily CPM Price for Display Ads Across Exchanges Varies by Over 100%
The fluctuation of ad CPM prices outpaced other dynamic values — even during a period that included the stock market’s “flash crash” and the Goldman Sachs’ investigation.
DataXu’s MarketPulse explores the data that defines today’s digital advertising marketplace. Stay tuned for future insights from DataXu. For a pdf of this report see DataXu MarketPulse: Ad Price Volatility, May 2010 and to sign up for future MarketPulse newsletters, send an email to marketpulse@dataxu.com.
Share on Facebook- Posted by DataXu on May 21st, 2010
DataView: The DataXu Blog
