The Path to a New Car: A Word from eXelate’s VP of Auto Sales

This is my first blog post as the automotive guy at eXelate, and being born and raised in Detroit, it’s been all about cars my entire life.  Now that data has become a major factor in targeting automotive shoppers, it just made sense for me to marry my auto marketing experience with smart data.

The auto shopper is one of the most sought after people across the web.   For years, marketers have been trying to figure out who is in-market and who are just tire kickers.  Are they looking for new or used or are they still undecided? And how long will it be until they really get serious and actually pick a model they like enough to go to the dealership to take a test drive?  Is that today or 3 months from now?

This is not a linear process like traditional media has taught us.  The purchase funnel we used to talk about is more of a path to purchase that actually resembles fingerprint technology.   Fingerprints might all look the same at a glance, but each one is different.  Everyone’s vehicle buying situation is unique and constantly changing.  Decisions need to be made for an infinite number of reasons; time frames and patterns are impossible to decipher.

Think about it.  Let’s say it’s January and one person wants a bigger car because they just moved to Buffalo, New York from Jacksonville, Florida and think that driving in the snow will be a challenge.  And another person wants to make sure they have a vehicle with the correct towing capacity for that bigger camper they’ve been eyeing.  Both definite SUV shoppers.  One has an immediate need and the other is still in the planning stages.  Yet you need to make sure you are targeting both individuals throughout their decision making process.  That’s where data comes to the rescue.  People’s online habits can give us clues to make marketing decisions more effective.

While OEMs are looking to target new vehicle shoppers, the auto shopping population has almost 3x as many used vehicle shoppers compared to new.

US Auto Sales, 2012 (http://www.nada.org/NR/rdonlyres/1B512AC7-DCFC-472C-A854-6F5527931A2F/0/2013_NADA_Data_062813.pdf)

US Auto Sales 2012 (Source: NADA)

When consumers go to auto shopping and research sites, the sites are designed to reach both new and used shoppers, so almost ¾ of your targeting is to the used buying market.  What if you could use a data source that concentrated on new car buyers and over indexed the marketplace to make targeting more effective?  That’s what eXelate has been doing in the auto vertical and that’s why the data is much more effective.

A recent 90 day Korrelate study found that eXelate data reaches half of all new car buyers.   In 2013 alone, you can reach 8 million auto buyers and 34 million that are in consideration.

This is an exciting time to be in the data business, and the auto industry understands how important it is to reach people throughout the entire path to purchase – wherever that road is to a new car sale.

Frank J. Smychynsky is VP of Automotive Sales at eXelate and is based in Detroit, MI. With over 20 years in the world of cars, Frank is our resident auto expert. Click here to get in touch with him. 

The Lyons Den: How Do You Measure Data Accuracy?

kevin lyonsThis week, eXelate unveiled the second white paper in our Smart Data series, accurate data is smart data. We believe that Big Data has yet to fulfill its promise of providing a clear and consistent business advantage. To address this, we argue that the industry needs to evolve from Big Data to Smart Data – data that drives business value because it is accurate, actionable, and agile. The accuracy white paper answers why accuracy is important, how to gauge it, and finally, what it takes to implement data accuracy.

exelate smart data

Which brings us to today’s topic, which builds on the white paper: How do you measure accuracy?

In measuring accuracy, it seems to me that many confuse accuracy and precision. Some, for example, might argue that a consensus approach which polls data providers is the best way to judge data accuracy. This approach essentially claims that attributes with the highest consensus across data providers is the most accurate. We categorically reject this simplistic view as it equates agreement – or precision – with accuracy. In fact, checking multiple data sources against one another may create a kind of confirmation bias (the tendency for people to believe data that supports what they already believe to be true).

This can happen when multiple vendors have the same technique to collect data. Take income data, for example. If several data vendors are supplying the mean income for a zip code to individual users, each will agree for a given household, but each will be wrong for most of the households. And, interestingly, if five data vendors applied the zip code mean to a specific user and one supplied the actual income to the same user, the actual income would be discounted as inaccurate. Serial correlation in data streams is often misinterpreted as accuracy. That is, everyone agreeing doesn’t necessarily make it right; it may just be that many data sources are wrong for the same reason – and that is why the consensus approach breaks down. Precision is not accuracy.

We therefore believe that online data must be validated with gold-standard, independent third party sources, and that these sources must contain registered (user-level) audience demography. Validating eXelate data against third party sources such as comScore and Nielsen allows us to verify that our data achieves the greatest possible balance of scale and accuracy, without degrading our data by including overly loose criteria.

There are a couple of very straightforward ways to validate the accuracy of data against a gold standard. For binary (either/or) outcomes, a confusion matrix is probably the most common approach. Let’s look at the simple case of gender. The below chart represents a simplified confusion matrix which allows us to judge the accuracy of our data:

equation1

The above chart reads as follows.  The rows represent what eXelate believes about a user. The columns are the “gold standard” against which we are being evaluated. So, from the perspective of correctly identifying males, there are four potential outcomes:

  • Accurate (male) – technically known as “true positive” – means that both eXelate and the “gold standard” knows the user to be male; or, we got it right.
  • Accurate (female) –  or “true negative” – means that both eXelate and the “gold standard” knows the user to be female,  so again we got it right (it’s a “true negative” because from the perspective of identifying males, we’re technically agreeing that this user is not male).
  • Inaccurate (actually male) – or “false negative” – means that eXelate believes the user to be female, but they are actually male; or, we identified this user incorrectly.
  • Inaccurate (actually female) – or “false positive” – means that eXelate believes the user to be male, but they are actually female; wrong again.

In the above, accuracy is defined as,

equation2

So, if you saw the following results:

equation3

your accuracy would be:

equation4

meaning that, overall, you were right 81% of the time.

To underscore the fact that precision is not equal to accuracy, even in technical terms, we can note that precision for males is defined as:

equation5

meaning that if you showed an ad to users that you thought were males, you would be on target  83% of the time.

These equations, as well as something called recall (of males, how many did we reach?), are KPIs eXelate employs to continuously evaluate and improve our data.

So, to summarize, accuracy matters!  Everyone in our business needs to understand what it is and what it is not. Accuracy needs to be calculated using accepted methods against gold standards. And those that fail to give it the attention it deserves do so at their own risk!

We’ll have much more to say on Smart Data and I encourage you to read our Smart Data series of white papers and continue to follow us @eXelate.

Audience Targeting Budgets to Increase by 38%: eXelate and Digiday’s State of the Industry

In March and April of 2013, we again participated in the State of the Industry Survey in conjunction with Digiday. The survey captured feedback from over 650 digital advertising professionals on the importance of audience targeting in launching high-performance direct response and branding digital campaigns. Respondents represented a wide range of industry stakeholders – advertisers, agencies, ad networks, ad exchanges, and demand side platforms (DSPs). The research was driven by the ongoing need to reduce the complexity of the ad-tech ecosystem by providing clarity to the real demands of advertisers and agencies.

digiday exelate soti

A surprising survey insight included the fact that both advertisers and agencies preferred 3rd party online data as the #1 source of audience targeting data over 1st party social media, 1st party custom, and 3rd party offline data for both direct response and branding campaigns. This result confirms that the audience targeting ecosystem continues to rely on accurate and actionable 3rd party data pools to power effective audience targeting campaigns.

Key highlights from the eXelate and Digiday State of the Industry survey include:

  • Audience targeting continues to grow with more than 80% of advertisers and 90% of agencies, ad networks, ad exchanges and DSPs utilizing the capability
  • Over 60% of advertisers prefer 3rd party online data for audience targeting, followed by 1st-party CRM data
  • Over 80% of the ecosystem reports that audience targeting is an effective marketing strategy
  • 3rd party online data is the highest ranked data set for both direct response and branding campaigns for both advertisers and agencies, outperforming 1st-party social, custom and CRM data, as well as 3rd-party offline data
  • More than 69% of respondents report they plan to increase their audience-targeting budgets, with an average budget increase of 38%

“Despite the ongoing debate on 3rd party cookies and data, it is clear that all parts of the digital advertising ecosystem – advertisers, agencies and platforms, increasingly rely on, and prefer the efficacy and reach of, 3rd party data,” stated Khurrum Malik, CMO (eXelate).

See the presentation:

Click here to view the full results.

eXelate CEO Mark Zagorski talks Data Modeling with ExchangeWire

As seen in: ExchangeWire

Sophisticated data modeling has become a requisite for most traders working in the data-driven ad space. Some are building their own prop models, others are partnering with the likes of eXelate to personalise the message for the audiences they are trying to reach. In this episode of #TraderTalkTV we get Mark Zagorski, CEO of Exelate, to explain his company’s approach to Real-Time Data Modeling.

Big Data Myths: Why All Data is NOT Created Equal

At Digiday Exchange in NYC in August, eXelate CEO eXelate CEO Mark Zagorski took to the stage to discuss Big Data, how companies should approach it, and some of the most common myths surrounding it. Take a look at the full interview below, or scroll down to read some of the discussion highlights.

 

On Big Data: Big Data isn’t about having a big database – it’s about having many pieces of unstructured data in which you look for signals that create patterns of consumer behavior upon which you can act. It’s not CRM data, but it’s about seeing traffic in and out of stores, knowing sales, and making connections between the two. Most companies think they need an all-encompassing solution for data. But in reality, there are a lot of small things brands and advertisers can do to claw away at each pile of data, and these small things can prove incremental results – for example, bringing your offline data online. The trick is to look at the specific thing you want to do – drive higher conversion rates or acquire more customers – and apply a specific data set to your problem.

Data Myth #1 – All data is created equal. There is good data, and there is bad data. There are valuable and non-valuable signals. When you look at the Lumascape (another myth!) there are a lot of different data companies. Every logo is the same size but in reality, every data set is not the same size nor do they have the same caliber of data. Take online data versus offline data; a modeled offline user is different than a registration-based online user when it comes to something like demographic information.

Data Myth #2 – First party data is more reliable than third party data. First party data is great – for the first 50 people that you’ll reach with it. The issue with first party data is that it doesn’t scale, and using it is essentially talking to the same customers that you’re already in contact with. That’s great for upselling and sales retention, but what about acquiring new customers? The best use for first party data is as a seed to model against – you want the best first party data to model a top-of-the-funnel audience.

Data Myth #3 – The Lumascape is an accurate depiction of the complexity of the ad tech world. All due respect to Terry Kawaja & Co, but the chart was created for one reason – to sell banking services. There are two key issues with the amount of logos on the slide – (1) there are a lot of opportunities for consolidation and (2) they create a lot of confusion, which in turn creates a market for services that will clear up that confusion. An advertiser or a publisher does not need to work with every single company on the Lumascape chain to make a connection – many times they’re not involved with any major decisions. The Lumascape overcomplicates areas that are simple, and oversimplifies some areas that are complicated.

Join the conversation – follow Mark and Digiday on Twitter.

Audience Data Quality Control – 4 Tips to Help Marketers Navigate Third Party Data Providers

In the last three years, integrating audience data to improve the targeting for online ad campaigns has become an industry standard. Unfortunately, not all audience data is created (and sourced) equally.

On July 27th, in an article about Google and 3rd party data, Laurie Sullivan interviewed Colette Dill-Lerner, the VP of Internet Marketing at leading direct marketing company Guthy-Renker, who said that when she reviewed data files, she found nearly 50% of their gender data was wrong. In essence, the data used provided as much value as a coin toss.

In the same way that a few spammers hurt the entire email marketing industry, providers of inaccurate data hurt the data targeting industry.

So how can marketers ensure that they’re getting quality data to enhance targeting for their online advertising campaigns? Here are a few tips for conducting your own audience data quality control:

  • Buy data which has been verified by an independent 3rd Party. In 1914, the Audit Bureau of Circulations was established to verify magazine circulations by advertisers, agencies and publishers in order to end deceptive pricing practices in the magazine industry. Today, most advertising vehicles have their reach/frequency/circulation data verified by an external body. So why isn’t audience data verified? The same companies which verify online traffic data also have the tools to verify audience data to ensure that the gender data is correct. Seek certification when you can.
  • Buy data from a data provider which purchased the data directly from the source. With the growth in audience data usage, more and more vendors have started incorporating audience data into their offerings. To ensure the quality of the data they purchase, agencies and marketers should buy data either directly from the source OR from a data provider who purchased the data directly from the source. This way, there will always be someone accountable for the data. You shouldn’t have to be a detective to sniff out a source.
  • Buy data from a provider with enough range, volume and activity to support your campaign goals and enough of a track record to vet the data. When buying audience data, the audience segment you think will perform best is not always the one which actually performs best. The best results come from testing and analyzing campaign performance, and optimizing the data segments as the campaign progresses. That’s why it’s important to purchase audience data from a provider with a broad enough range of data segments to enable testing and analysis, and enough scale to reach your campaign goals. Plus, a long history of successfully performing campaigns at scale provides feedback to ensure that segments are regularly vetted and put to the test in the real world.
  • Make sure that your verification methods match the data capture and targeting methods. If you are buying data that is derived at a household level and your campaign goals are targeting at an individual level, you have a data/performance mismatch that will not only result in poor verification scores, but also poor campaign performance. Ensure that your measurement criteria match with how the data was sourced and what your targeting efforts are trying to accomplish. A great place to start to understand data definitions is the IAB’s Data Lexicon, a detailed primer on the who, what and where of data sourcing. Additionally, it is critical to “test your tester” – dig deep into the methodology and check that it too complies with industry standards. As accuracy standards still aren’t “standard”, incompatibility between data collection and verification processes are rampant, and can unfortunately provide both false positives and negatives.

Audience targeting data is driving huge gains in campaign ROI. But, as its use proliferates, agencies and marketers need to take the necessary quality control measures to ensure their data sources are transparent, the data product is regularly vetted at scale and that campaign goals and performance metrics are aligned.

Modeling eXposed – A Video Recap

Last Wednesday, eXelate’s modeling and analytics event, Modeling eXposed, addressed how innovations in data modeling can be implemented into your digital marketing programs for more effective targeting, campaign management, business intelligence, and analytics. The purpose of the event was to bring industry peers together to share knowledge and exchange ideas for improving advertising performance through flexible modeling solutions.

Through a series of presentations and discussions, eXelate and guest speakers addressed innovations in modeling through case studies and interactive panel discussion, providing attendees with a roadmap of how modeling can be successfully implemented into their marketing strategies. Topics included:

  • Data analytics for improved marketing decisions
  • Custom audiences for cross-channel media optimization
  • Adaptive audience intelligence methodology
  • Modeling as a direct response and customer relationship strategy

Check out our recap video below. For more pictures and information, visit www.modelingexposed.com.

Sh*t Digital Media Folks Say – Brought to you by eXelate

Wanna meet at Ace Hotel?

It’s no wonder why most people refer to the advertising technology space as an “ecosystem.” We are all in the same sphere – we attend the same conferences, know the same people, and use the same lingo. While this can be humorous, it’s also the driving force behind our success. Our industry’s close-knit community is the reason why we are able to come together, collaborate and propel fantastic innovation.

We teamed up with Meerkat Media to try and recreate in 3 minutes what most digital media “folks” hear every day and every week at work. For your viewing pleasure – Sh*t Digital Media Folks Say.

Special thanks to:

Joe Apprendi

Joanna O’Connell

Lauren Adams

Dru Johnston

And of course, some of our hard-working eXelate team – data skills AND acting skills!

  • Mark Zagorski
  • Rudy Feliciano
  • Vlada Kabatyanskaya
  • Dale Campbell
  • Emily Spence
  • Julie Pereira
  • Meghan Brown

Q&A with New eXelate Partner TruSignal

eXelate announced last week that we partnered with TruSignal in our continued effort to bring high-quality audience targeting to marketers. TruSignal’s suite of TruAudience℠ High Value Consumer segments are now publicly available to all of our integrated media platforms. In an effort to introduce you to our latest partner, we had a few questions  for TruSignal’s CRO Jeremy Longinotti.

Take a look and let us know what value you think this partnership will bring to the digital marketing world!

What does this partnership with eXelate mean for TruSignal?

We’ve been really happy with the team at eXelate and their expertise in the digital ad tech ecosystem. As the relationship grows, we are hoping to see an increase in use of and interest in our TruAudience Syndicated Segments. Because of our unique approach, our data segments can be used in both branding and direct marketing campaigns and eXelate will continue to educate the market on the best ways to use data in online targeting campaigns. We also look forward to more innovation from eXelate as they continue to build out their platforms and provide more ways for buyers and sellers to connect.

TruSignal has a unique approach to building audiences, can you explain that?

TruSignal’s approach takes first party customer data and combines it with our own massive database of offline consumer-profile data. With predictive modeling, we find the right combinations of data attributes that generate the most predictive signal for our syndicated segments.  What’s unique about this?  It all happens in an offline manner, distilling all of the data into a single, segmented audience formula. We’re not talking inferred or clickstream data, we’re using real life data—historical performance and consumer profile data from 40 data sources. By going back to the raw offline data, we can access a huge set of over 10,000 attributes, about 80% of which have not been onboarded into the online ecosystem. The result is an audience formula that finds “lookalike” prospects with the highest propensity for a product or service, specific to each of our nine TruAudience syndicated segments. The following TruAudience syndicated data segments are now available through eXelate:

    • Auto Insurance Online Propensity
    • Term Life Insurance Propensity
    • Online Higher Education Propensity
    • Mortgage Refinance Propensity
    • Political Donor Propensity (Democrat & Republican)
    • Political Affiliation
    • Estimated Financial Health
    • Estimated Household Income
    • Underbanked Consumers

Given the importance of your third party data in the process, can you tell us a little more about the data you use to build your vertical specific syndicated audiences?

TruSignal licenses and curates consumer profile data on almost every U.S. household and adult from over 40 high-quality, stable and verified data sources. What this means is that we have real life data such as summarized financial information, census data, public records, lifestyles, catalog purchase activity, hobbies/interests and demographics. From this data, we access over 10,000 predictive attributes which are used in combination with actual customer or industry-specific performance data. Another great part about the data is that we do all the heavy lifting for you. We not only use a proven regression technique that determines which data attributes are the right ones for the segments, but our data team is consistently testing each of our data sources to ensure our data is the most accurate, comprehensive, reliable and predictive data available.

For more information on our partnership with TruSignal, click here.

Combining Data to Create Powerful Modeling

The Netflix Tech Blog recently posted an entry that discusses their recommendation algorithms and outlined the Netflix Prize, a machine learning and data mining competition to predict movie ratings. The 2009 winner of the contest improved Netflix’s ratings prediction system by more than 10% with a new algorithm. For a company owing 75% of viewership to recommendations, this would seem to be a huge step for Netflix. They didn’t adopt the winner of the contest, however – and the reasoning is perfectly logical.

The winning algorithm focused mainly on predicting ratings. Ratings are an important source of data for Netflix, but new types of analyses & inputs beyond ratings alone have emerged, all of which can help Netflix create even better recommendations. These include context, movie title popularity, novelty, diversity and freshness.

If you’ve been reading this blog or our Twitter handle for the last 2 months, then you are probably familiar with maX™ – or Modeled Audience eXtension, our custom modeling product that helps advertisers build a scaled audience with broad reach that is 3-5x more likely to respond to their campaigns.

We make it work by focusing on all the data – the advertiser’s first party data as well our eXelate premium marketplace data. And we look at as many data points as possible, including demographic, lifestyle, intent, behavior, brand affinity information and other proprietary data sets many advertisers may not have access to. We’ve learned, as Netflix seems to have, that combining multiple forms of data can prove to be the most powerful way to reach an audience. If you were Netflix, what would you have done?