Kelley Blue Book

Increases ad revenue with better, faster data analysis and ad price optimization

Published on 26-Jul-2012

"Data from our Netezza-powered Advertising Reporting Tool and from Search Ignite [SEM platform] are now in alignment so that we consistently deliver ‘committed’ impressions. Our inventory forecasting accuracy is up, which has reduced our penalties for shortfalls." - Karen Simmons, Senior Director of Data Warehousing, Kelley Blue Book

Customer:
Kelley Blue Book

Industry:
Automotive, Retail

Deployment country:
United States

Solution:
BA - Business Intelligence, Data Warehouse, Information Management Foundation, Smarter Computing, Smarter Planet

Smarter Planet:
Smarter Solutions for Retail

Overview

In the early 1920s, the owner of Kelley Kar Company in Los Angeles began distributing to other dealers and banks a list of automobiles he wanted to buy and the prices he would pay. In 1926, Kelley’s inventory wish list became the Blue Book of Motor Car Values, offering factory list prices and cash values for thousands of vehicles, and establishing “Blue Book value” as the standard of car valuation referenced by sellers and buyers.

Business need:
Advertising data volumes exceeded the capability of the existing SQL Server environment, slowing data loading and queries. The company also needed greater computational power to estimate advertising inventory availability a year in advance.

Solution:
With initial deployment completed in two days, the IBM® Netezza data warehouse appliance provides the load and scale to support the company’s analytics requirements.

Benefits:
Increased ad revenue from better ad analysis, forecasting and fill rate; improved profitability from ad pricing and search engine marketing optimization; strengthened customer satisfaction with faster, accurate publication of auto valuations.

Case Study

In the early 1920s, the owner of Kelley Kar Company in Los Angeles began distributing to other dealers and banks a list of automobiles he wanted to buy and the prices he would pay. Demand grew for “Kelley’s Cash Price List” as the trade recognized Les Kelley’s ability to predict market values accurately. In 1926, Kelley’s inventory wish list became the Blue Book of Motor Car Values, offering factory list prices and cash values for thousands of vehicles, and establishing “Blue Book value” as the standard of car valuation referenced by sellers and buyers.

More than 80 years later, Kelley Blue Book (KBB) is still innovating and transforming itself. The publisher that has long provided a reliable mechanism for buying and selling cars is rapidly becoming an analytics- driven information powerhouse that facilitates the buying process for consumers while providing qualified leads and market insights to dealers and OEMs (original equipment manufacturers). The company estimates that each month its website, KBB.com, experiences more than 16 million visits making it a top online destination for car shoppers. Because three of every four visitors are “undecided,” KBB.com is also a “must buy” on the advertising media plans of every OEM. Car dealerships count on a steady flow of leads from consumers who generate more than 30 million pricing reports monthly.

Analytics: the heart of KBB’s strategy
From “Kelley’s Cash Price List” through today, KBB at its core has always been a data company, but until three years ago, analytics played a minor role in most decisions. That is changing, according to Dan Ingle, Vice President of Analytic Insights: “Now analytics are at the heart of KBB’s strategy. We use analytics to optimize lead generation, and analytics enable us to maintain a very low error rate in forecasting vehicle valuation. With Netezza [now IBM ], we’re able to process all of our forecast models in a day, compared with the previous three to four days. This enables us to produce vehicle values that we can deliver in near real-time to the marketplace instead of waiting up to two weeks to push those values out to KBB.com.”

It is in online advertising sales, however, where analytics capabilities arguably have become the most important differentiator available to websites in highly competitive ad-driven markets. The fractured, fast-growing online ad business still depends largely on ad server technology and workflow conceived in the 1990s to solve simpler problems than today’s data-intensive riddles, such as inventory forecasting, impression and page revenue yield management, audience- based targeting and ad price optimization.

Like ad server technology designed for less complex tasks, KBB’s existing Microsoft® SQL Server® environment also proved to be inadequate. Advertising data volumes would swell to 10 terabytes, and data loads and queries were taking too long. Other critical business centers faced similar performance challenges; for example, the legacy system used to set vehicle values was antiquated. The time had come to build a data infrastructure that could handle the current load and scale with KBB’s aggressive goals.

KBB decided to implement a business intelligence (BI) platform to improve reporting and visualization capabilities. However, with the BI platform pulling data from the SQL Server warehouse, the fastest queries would still take hours to days to complete, or would time out and never return at all. At the same time, these queries were taxing the server and negatively impacting other processes that needed to run. Many of the BI platform features, such as automated dashboarding capabilities, were rendered useless due to the processing time needed through SQL Server.

KBB knew that it was time to evaluate a new data warehouse platform to empower their BI capabilities. After evaluating the performance, total cost of ownership (TCO) and ease of implementation of data warehouse appliances, KBB decided to move forward with IBM Netezza data warehouse appliance. The appliance offered fast analytic performance, integrated seamlessly with KBB’s existing BI and SAS environments, and represented lower TCO than similar competitive products in terms of pricing, licensing and maintenance costs.

IBM Netezza data warehouse appliance was selected to help gather, analyze and distribute data from many different sources much faster and in multiple applications. The logical first application that KBB deployed in production on the platform was their Advertising Reporting Tool (ART), which was up and running in a matter of days. “Netezza is one of the best companies I’ve ever worked with – a true cohesive relationship,” says Karen Simmons, Senior Director of Data Warehousing at KBB.

“They delivered on what they said they would, got us up and running, and then they stayed after to make sure the team understood how to use it and how we could do special things with the data.”

Once the IBM Netezza data warehouse appliance was deployed, KBB found they could take advantage of BI features that were previously inaccessible due to performance issues. All of a sudden, queries that
took days came back in minutes. With SQL Server, the more KBB tried to narrow down their data, the slower it would come back. With the IBM Netezza data warehouse appliance, the opposite is true: the more filters used, the faster the performance. Kelley Blue Book built out the ART system to allow granular access to the data, so the ability to dive into detailed data quickly made the IBM Netezza data warehouse appliance a huge success. The IBM Netezza appliance is going to allow KBB to combine advertising data, revenue data, and site traffic data – they can merge several different data sources and get a multi-channel view of their business.

ART: accuracy and speed provide path to higher ad revenue
As a result of the legacy of television advertising buying schedules, auto makers traditionally plan and buy much of their online advertising according to an “upfront” schedule. For Kelley Blue Book, this means using historical ad and site data – site visits, page views, historical trends and granular ad impression history – as early as March to forecast and allocate advertising inventory for each OEM for the next 12 to 24 months. There are consequences for miscalculating; if KBB delivers more ads than estimated, they receive no additional revenue for the overage. If they deliver fewer impressions than estimated, KBB returns money to the advertiser. In both cases, poor forecasting can hurt credibility and the relationship with the client. The IBM Netezza data warehouse appliance helps Kelley Blue Book mitigate financial risks.

Impressions analysis improves forecasting capabilities
KBB experienced its first big win with the IBM Netezza data warehouse appliance after migrating impressions data to the appliance, which provided better forecasting capabilities. Before, extracting impression, click and rich media data from Kelley Blue Book’s DoubleClick DART ad server into SQL Server required one to two weeks and much tuning. A high level SAS optimization analysis that would take one or two days in SQL Server now runs on the IBM Netezza data warehouse appliance in one or two minutes. Queries that took SQL Server an hour take just three seconds on the IBM Netezza data warehouse appliance. With the IBM Netezza appliance, KBB can update its forecasts faster, using more data and resulting in higher accuracy.

Because so many of KBB’s ad revenues are sold up front, having the ability to create accurate and reliable impression forecasts is helpful in generating ad revenues and establishing trust with clients. “Now our clients are always telling us that our forecasts are among the best within our competitors,” says Simmons. “We’ve had auto makers and competitors asking for insight into how we’re able to forecast as accurately as we do.”

According to Ingle, “The company’s ability to pinpoint estimated impressions is worth millions to KBB. We have much more flexibility with Netezza. More of our analysts have access to more accurate data than ever. Our models now generate accurate forecasts quickly for OEMs that literally would have taken months before.”

The integration of SAS and the IBM Netezza data warehouse appliance is allowing Kelley Blue Book to automate much of the forecasting process.“Upfront forecasting is not a one-time effort each year,” says Simmons. “As a result of different clients’ timing, it is an almost nine-month long project. Being able to automate the update of the models frees up resources for other projects.” In fact, Simmons noted that with the ability to gather, analyze and distribute data faster from many different sources, there is now a clear path to more efficient site monetization: “Netezza is a critical component in the tech stack that we use to analyze our DART data and generate more ad revenue using existing data,” says Simmons. “It is one of the best investments we have made within our database infrastructure. More accurate analysis of ads enables us to increase the ad ‘fill’ rate, which translates into higher revenues for KBB.”

Simmons cites an example of a negotiation with an OEM: “Let’s say our sales executive is working with an OEM. She asks us how much inventory is available for that OEM and how and where it should be spent to maximize their investment and our returns. Before, we had to pull the data manually and painstakingly compare it with its competitors, one by one. Thanks to Netezza, now we can compare that OEM optimized inventory with all of its competitors and make more informed decisions much faster than ever before.”

After migrating impressions data to IBM Netezza data warehouse appliance, KBB’s next steps were to get click data and rich media data onto the data warehouse appliance. By analyzing click data, KBB can evaluate its ad offerings based on where the ads run and perform in different areas of the site. It can now compare the operational costs of running and managing each ad to the revenue generated by that specific ad to determine whether or not it is worthwhile to sell that particular placement. This is a strategic application for KBB that is helping it to increase overall revenues and improve trust in KBB ad forecasting.

“The ability to analyze our product offerings and look at how they perform across the site is invaluable in improving them,” says Simmons. “We have been collecting this performance data since 2008, but had enough performance problems with SQL Server when it was dealing only with impressions data; adding click or rich media data was not feasible. In fact, many at KBB find it hard to believe we didn’t have access to these insights all along, because they are so crucial to understanding our ad product.”

KBB can also perform iterative analyses that lead to smarter ad pricing. As opposed to using typical rate card pricing – so much for so many impressions of this type in this content – KBB is moving to a more dynamic system in which the data will help the company price impressions based on their value to the OEM. Having access to detailed data over long histories is allowing KBB to analyze the offerings it has, identify their value, and be in a better position to negotiate rate cards. The company is able to use analytics to better negotiate pricing with OEMs, based on better insight into the value of offerings. This sets the stage for individualized, dynamic pricing.

Once KBB finishes migrating rich media data to the IBM Netezza data warehouse appliance, it will have the capability to see how people are interacting with the ads. For example, it will be able to monitor whether people are expanding the ad window, where they’re hovering, how long they’re viewing video ads, and more.

Next up: SEM optimization
Now that Kelley Blue Book has migrated ad and valuations data to the IBM Netezza data warehouse appliance, the next step is to integrate web analytics and search engine management (SEM) data. This will offer KBB the ability to optimize SEM spending and align it correctly with the needs of ad sales for the right delivery of impressions to the right consumers. With the help of the IBM Netezza data warehouse appliance, KBB will be able to filter data to Search Ignite to optimize keyword performance, a capability that will sharpen its aim, maximize its search investment, and help it deliver content that improves the consumer’s overall experience on the site. This will give consumers greater confidence and comfort with the process of researching and buying a new or used car.

Paying off today, and opening the door to a data-driven future
The initial deployment of IBM Netezza data warehouse appliance took just two days. In its first year in production, the appliance is delivering performance, results and benefits appreciated by KBB’s consumers, dealers, OEMs, and employees. Simmons comments, “Our CEO and product managers are very happy with the interactive dashboards that they can now access directly through MicroStrategy. Product managers can run many different pre-built dashboards with drill down capabilities without requiring an analytic resource. Our analysts can focus on actually analyzing and providing insights, rather than pulling together information for a report. I don’t think I’m the only one wearing my ‘I heart Netezza’ shirt!” Here are some of the other benefits KBB is realizing from its use of the IBM Netezza data warehouse appliance:

• More ad revenue resulting from
– more accurate and speedier ad analysis
– improved ad inventory forecasting and fill rate

• Increased customer satisfaction resulting from
– faster valuations (model processing reduced from three days to one)
– alignment of ad and consumer data to make decisions leading to
delivery of better overall experience on the website

• More profitable operations resulting from
– optimization of ad pricing
– reallocation of valuable staff time (only one part-time DBA
needed to manage the environment)

These are still the early days of KBB as an analytics-driven organization. According to Ingle, “Analytics will be a primary source of our competitive advantage – and Netezza is the foundation of our data-driven strategy.”

About IBM Netezza data warehouse appliances IBM Netezza data warehouse appliances revolutionized data warehousing and advanced analytics by integrating database, server and storage into a single, easy-to-manage appliance that requires minimal set-up and ongoing administration while producing faster and more consistent analytic performance. The IBM Netezza data warehouse appliance family simplifies business analytics dramatically by consolidating all analytic activity in the appliance, right where the data resides, for industry-leading performance. Visit ibm.com/ software/data/netezza to see how our family of data warehouse
appliances eliminates complexity at every step and helps you drive true business value for your organization. For the latest data warehouse and advanced analytics blogs, videos and more, please visit: thinking. netezza.com.

About IBM Data Warehousing and Analytics

Solutions
IBM provides the broadest and most comprehensive portfolio of data warehousing, information management and business analytic software, hardware and solutions to help customers maximize the value of their information assets and discover new insights to make better and faster decisions and optimize their business outcomes.

For more information
To learn more about the IBM Data Warehousing and Analytics Solutions, please contact your IBM sales representative or IBM Business Partner or visit: ibm.com/software/data/netezza

Products and services used

IBM products and services that were used in this case study.

Software:
IBM Netezza Analytics

Legal Information

© Copyright IBM Corporation 2012 IBM Corporation Software Group Route 100 Somers, NY 10589 Produced in the United States July 2012 IBM, the IBM logo and ibm.com are trademarks of International Business Machines Corp., registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the Web at “Copyright and trademark information” at ibm.com/legal/copytrade.shtml Netezza® is a trademark or registered trademark of IBM International Group B.V., an IBM Company. This document is current as of the initial date of publication and may be changed by IBM at any time. Not all offerings are available in every country in which IBM operates. The performance data and client examples cited are presented for illustrative purposes only. Actual performance results may vary depending on specific configurations and operating conditions. THE INFORMATION IN THIS DOCUMENT IS PROVIDED “AS IS” WITHOUT ANY WARRANTY, EXPRESS OR IMPLIED, INCLUDING WITHOUT ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND ANY WARRANTY OR CONDITION OF NON-INFRINGEMENT. IBM products are warranted according to the terms and conditions of the agreements under which they are provided.