
In this article…
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Phasellus at nisl nunc. Sed et nunc a erat vestibulum faucibus. Sed fermentum placerat mi aliquet vulputate. In hac habitasse platea dictumst. Maecenas ante dolor, venenatis vitae neque pulvinar, gravida gravida quam. Phasellus tempor rhoncus ante, ac viverra justo scelerisque at. Sed sollicitudin elit vitae est lobortis luctus. Mauris vel ex at metus cursus vestibulum lobortis cursus quam. Donec egestas cursus ex quis molestie. Mauris vel porttitor sapien. Curabitur tempor velit nulla, in tempor enim lacinia vitae. Sed cursus nunc nec auctor aliquam. Morbi fermentum, nisl nec pulvinar dapibus, lectus justo commodo lectus, eu interdum dolor metus et risus. Vivamus bibendum dolor tellus, ut efficitur nibh porttitor nec.
Pellentesque habitant morbi tristique senectus et netus et malesuada fames ac turpis egestas. Maecenas facilisis pellentesque urna, et porta risus ornare id. Morbi augue sem, finibus quis turpis vitae, lobortis malesuada erat. Nullam vehicula rutrum urna et rutrum. Mauris convallis ac quam eget ornare. Nunc pellentesque risus dapibus nibh auctor tempor. Nulla neque tortor, feugiat in aliquet eget, tempus eget justo. Praesent vehicula aliquet tellus, ac bibendum tortor ullamcorper sit amet. Pellentesque tempus lacus eget aliquet euismod. Nam quis sapien metus. Nam eu interdum orci. Sed consequat, lectus quis interdum placerat, purus leo venenatis mi, ut ullamcorper dui lorem sit amet nunc. Donec semper suscipit quam eu blandit. Sed quis maximus metus. Nullam efficitur efficitur viverra. Curabitur egestas eu arcu in cursus.
H1
H2
H3
H4
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Vestibulum dapibus ullamcorper ex, sed congue massa. Duis at fringilla nisi. Aenean eu nibh vitae quam auctor ultrices. Donec consequat mattis viverra. Morbi sed egestas ante. Vivamus ornare nulla sapien. Integer mollis semper egestas. Cras vehicula erat eu ligula commodo vestibulum. Fusce at pulvinar urna, ut iaculis eros. Pellentesque volutpat leo non dui aliquet, sagittis auctor tellus accumsan. Curabitur nibh mauris, placerat sed pulvinar in, ullamcorper non nunc. Praesent id imperdiet lorem.
H5
Curabitur id purus est. Fusce porttitor tortor ut ante volutpat egestas. Quisque imperdiet lobortis justo, ac vulputate eros imperdiet ut. Phasellus erat urna, pulvinar id turpis sit amet, aliquet dictum metus. Fusce et dapibus ipsum, at lacinia purus. Vestibulum euismod lectus quis ex porta, eget elementum elit fermentum. Sed semper convallis urna, at ultrices nibh euismod eu. Cras ultrices sem quis arcu fermentum viverra. Nullam hendrerit venenatis orci, id dictum leo elementum et. Sed mattis facilisis lectus ac laoreet. Nam a turpis mattis, egestas augue eu, faucibus ex. Integer pulvinar ut risus id auctor. Sed in mauris convallis, interdum mi non, sodales lorem. Praesent dignissim libero ligula, eu mattis nibh convallis a. Nunc pulvinar venenatis leo, ac rhoncus eros euismod sed. Quisque vulputate faucibus elit, vitae varius arcu congue et.
Ut convallis cursus dictum. In hac habitasse platea dictumst. Ut eleifend eget erat vitae tempor. Nam tempus pulvinar dui, ac auctor augue pharetra nec. Sed magna augue, interdum a gravida ac, lacinia quis erat. Pellentesque fermentum in enim at tempor. Proin suscipit, odio ut lobortis semper, est dolor maximus elit, ac fringilla lorem ex eu mauris.
- Phasellus vitae elit et dui fermentum ornare. Vestibulum non odio nec nulla accumsan feugiat nec eu nibh. Cras tincidunt sem sed lacinia mollis. Vivamus augue justo, placerat vel euismod vitae, feugiat at sapien. Maecenas sed blandit dolor. Maecenas vel mauris arcu. Morbi id ligula congue, feugiat nisl nec, vulputate purus. Nunc nec aliquet tortor. Maecenas interdum lectus a hendrerit tristique. Ut sit amet feugiat velit.
- Test
- Yes

As I mentioned in my previous blog, model validation is an essential step in evaluating a recently developed predictive model’s performance before finalizing and proceeding with implementation. An in-time validation sample is created to set aside a portion of the total model development sample so the predictive accuracy can be measured on a data sample not used to develop the model. However, if few records in the target performance group are available, splitting the total model development sample into the development and in-time validation samples will leave too few records in the target group for use during model development. An alternative approach to generating a validation sample is to use a resampling technique. There are many different types and variations of resampling methods. This blog will address a few common techniques. Jackknife technique — An iterative process whereby an observation is removed from each subsequent sample generation. So if there are N number of observations in the data, jackknifing calculates the model estimates on N – 1 different samples, with each sample having N – 1 observations. The model then is applied to each sample, and an average of the model predictions across all samples is derived to generate an overall measure of model performance and prediction accuracy. The jackknife technique can be broadened to a group of observations removed from each subsequent sample generation while giving equal opportunity for inclusion and exclusion to each observation in the data set. K-fold cross-validation — Generates multiple validation data sets from the holdout sample created for the model validation exercise, i.e., the holdout data is split into K subsets. The model then is applied to the K validation subsets, with each subset held out during the iterative process as the validation set while the model scores the remaining K-1 subsets. Again, an average of the predictions across the multiple validation samples is used to create an overall measure of model performance and prediction accuracy. Bootstrap technique — Generates subsets from the full model development data sample, with replacement, producing multiple samples generally of equal size. Thus, with a total sample size of N, this technique generates N random samples such that a single observation can be present in multiple subsets while another observation may not be present in any of the generated subsets. The generated samples are combined into a simulated larger data sample that then can be split into a development and an in-time, or holdout, validation sample. Before selecting a resampling technique, it’s important to check and verify data assumptions for each technique against the data sample selected for your model development, as some resampling techniques are more sensitive than others to violations of data assumptions. Learn more about how Experian Decision Analytics can help you with your custom model development.

There’s no question today’s consumers have high expectations. As financial services companies wrestle with the laws and consumer demands, here are a few points to consider: While digital delivery channels may be new, the underlying credit product remains the same. With digital delivery, adhere to credit regulations, but build in enhanced policies and technological protocols. Consult your legal, risk and compliance teams regularly. Embrace the multitude of delivery methods, including email, text, digital display and beyond. When using the latest technology, you need to work with the right partners. They can help you respect the data and consumer privacy laws, which is the foundation on which strategies should be built. Learn more

There are many factors attributing to the success of dealerships. When it comes to dealers, empirical guidance is a great way to study effective advertising. Experian brought Auto, Targeting, and the Dealer Positioning System capabilities together in a nationwide study to answer the ultimate question: what drives sales? The answers can be found in Experian’s 2018 Attribution Study. This is a wide-ranging, dealer-focused sales-driven attribution study that analyzed a few key variables. We deployed 187,701 tracking pixels to devices in 41,012 distinct households, focused on 15 digital metrics to learn about shopper behavior, and tied that digital shopping data to 2,436 vehicle sales. An industry first, Experian’s ability to combine automotive registration data, sales data, and website analytics and online behavior data puts us in a position to do something that very few companies can do. We use the household identifiers to not only see who bought a car and who bought specifically from a participating dealer, but also how they shopped the dealer’s site. Our ability to accurately identify a household’s digital behavior is based on the fact that we are a source compiler of the data and have it sitting under one roof. Others that attempt to provide this type of insight need to contract out for registrations, sales data imports from the dealership, website analytics, household identifiers, or all the above, which generally adds time to the insights. Using our sales-based approach, we can deliver unbiased attribution. Sales-based attribution is attributing credit to different advertising sources/campaigns based on actual vehicle sales – including those targeted consumers that may have purchased outside of the dealership. This is the Holy Grail of attribution for car dealers since it ties an offline activity such as buying a car back to the online advertising that’s taking up most their budgets every month. Because of that offline-online disconnect, sales-based attribution is difficult. Other automotive attribution models are typically focused on website conversions or website behavior – “what advertising can I attribute website leads to” (conversions) or “what advertising is driving users who follow the behavior that I think shows they’re likely to buy from my dealership” (website behavior.) What are the takeaways? We found three takeaways from our study. First off, we look at shopper behavior instead of isolating KPIs. Later we will discuss how traditional website metrics do not tie-in to sales. Second, we look at optimizing your paid advertising. Finally, we look at third-party investments. Although third parties drive sales, they may not be your sales. Looking at shopper behavior, not isolated KPI’s Traditional website metrics don’t tell the sales story for dealers. Traditional conversion stats are equal for buyers vs. all traffic such as VDPs or page views What this means is on average, buyers converted at a lower rate than overall website traffic. Looking solely at form submissions, hours and directions pageviews, and mobile clicks-to-call, don’t give the best view of what advertising is driving sales. With that, 98% of buyer traffic never submitted a form or went to the hours and directions page. This is a typical website conversion that dealers, vendors, and advertising agencies focus on. Since traditional web metrics don’t tell the story, there is another way. These are called High-Value Users, or HVU. They purchase at a 34% higher rate than overall traffic although they make up 11% of all traffic. High-Value Users are an Experian derived KPI. What makes someone an HVU are four different measurements. They must visit a website at least three times Spend at least six minutes on the site in total View at least eight pages in total View at least one VDP High-Value Users correlates to sales better than Vehicle Detail Page or VDP metrics. In this study, the correlation for VDP was measured at .595 which is rated a medium correlation. Meanwhile, HVU scored a .698 which is rated a high correlation. Looking at many different behavioral KPIs, like we do with our High-Value User (HVU) metric, correlates better to sales than just looking at how many VDPs you had. Driving more VDPs won’t necessarily help sales. But driving more HVUs is more likely to correlate with more sales. This also gets back to the attribution discussion above: Experian sales-based attribution is the best, and Experian’s HVUs are a good method for web-based attribution. From this attribution study, High-Value Users are a vital group for dealers to utilize. In our next post, we will go over the second and third takeaways from the attribution study: optimizing paid advertising and evaluating third-party investments.


