ellipsis flag icon-blogicon-check icon-comments icon-email icon-error icon-facebook icon-follow-comment icon-googleicon-hamburger icon-imedia-blog icon-imediaicon-instagramicon-left-arrow icon-linked-in icon-linked icon-linkedin icon-multi-page-view icon-person icon-print icon-right-arrow icon-save icon-searchicon-share-arrow icon-single-page-view icon-tag icon-twitter icon-unfollow icon-upload icon-valid icon-video-play icon-views icon-website icon-youtubelogo-imedia-white logo-imedia logo-mediaWhite review-star thumbs_down thumbs_up

Real-time bidding 101

Since its launch in 2009, real-time bidding (RTB) has revolutionized the display advertising industry. With impression-based bidding and targeting, RTB brings the advertising relevance and efficiency of the search auction to the world of display. Over time, we've seen more and more advertising dollars shift from traditional ad buys to RTB. If you're a marketer considering RTB for your campaigns, here are a few tips to get started.

Every customer is different

As an advertiser, you realize that each customer is slightly different from the next. As individual actions may vary, it may not always make sense to group customers into buckets as if they're all the same. For example, "business travelers" or "sports enthusiasts" can be vastly different individuals. Everyone in the world has a unique set of characteristics -- different income levels, hobbies, family statuses, etc. In addition, grouping customers into segments relies, in part, on human intuition, but the factors that drive performance can often not be intuitive.

Let's say you are running a campaign for an airline specific to a destination. Intuitively you would imagine that your ads would be running on sites that appeal to travelers, yet when you examine the sites where the campaign was serving ads, you notice that the ads are over-indexing to dentist-related sites. Upon further investigation you discover that there was an upcoming convention for dentists at the destination. If a human had managed the campaign, the dentist convention would most likely not have been found, and the human wouldn't have intuitively thought to serve ads on dentist-related sites.
RTB-powered advertising allows you to bid differently for each impression, therefore, presents a unique opportunity to target individuals -- segments of one. While this would be impossible for any human to do, this is a perfect application of algorithms and machine learning. Automated buying allows you to model exactly what's impacting your stated performance goal versus guessing what will work best for your campaigns.

Reaching and influencing customers at the right time is crucial

There are seemingly limitless data sources that advertisers can use for targeting consumers, so advertisers need to remember that not all data sources are created equal. Freshness is a common problem, and data can be out-of-date even before it reaches an advertiser. In fact, most data sets used in RTB are over a week old and may be too stale to identify the right customer at the right point in his or her purchasing journey. Additionally, few data sources are large enough on their own to provide meaningful targeting insights, so they are often combined in order to reach scale. Both a lack of freshness or data integrity can lead to data gaps, data inconsistencies, and flawed targeting models.
A truly effective display ad-targeting model requires fresh data on a massive scale.
Freshness is important because it gives you an accurate picture of where your customers are in the purchase cycle. It can also identify where in the purchasing funnel customers are so that marketers can provide them with the most relevant marketing messages possible. For example, if your average customer takes four days to make a purchase on your website, and your cookies are a week old, you're missing the entire purchase cycle for that customer, and as a result, the real opportunities to influence their path to purchase.
While necessary for some advertisers with shorter purchase cycles, however, not all advertisers will need data this up-to-date to create effective targeting models.

When choosing which type of data to use, you should evaluate your customers and their purchase cycles, and then seek out a first- or third-party data source with the level of freshness you need to build the best targeting model to reach your customers.

A consumer's path to purchase is not always linear

Retargeting has emerged as an effective tactic for bringing interested prospects back to your website to make a purchase. With the popularity of last-touch attribution -- where you give 100 percent of a credit for a conversion to the last ad touchpoint -- it's no surprise that retargeting spending has exponentially increased over the past number of years. However, there are a few reasons to look beyond this tactic for fulfilling your marketing needs. First, retargeting addresses only one part of the customer journey and prospecting efforts; those that bring customers to your site to begin with are a baseline requirement for growth but are typically not valued in last-touch attribution. Second, there are a number of dubious retargeters in the marketplace who game the system by bombarding site visitors with ads in an effort to capture attribution credit, regardless of whether those advertisements offer value or truly drove performance. The result of this is that the tactics that are truly driving performance become obfuscated and risk misdirecting future budget investments.

Instead of giving 100 percent of the credit for a conversion to the last click or view, you may want to consider a multi-touch attribution model that distributes credit across the conversion process. Adopting a multi-touch approach may reveal that your upper-funnel targeting is doing some of the heavy lifting for your conversions. If this is the case, you'll want to strike a balance between your prospecting and retargeting spend. Prospecting is important for all advertisers and helps refill your marketing funnel after customers have converted.


Display advertising is only as effective as what you put into it. When you use the wrong type of data, targeting methods, and attribution modeling, you end up with less than stellar advertising results. On the other hand, when you use large, fresh data sets, individual targeting capabilities, and multi-touch attribution modeling, you can see that quality inputs transform directly into new prospects, repeat site visitors, and conversions. Try it out and see the increased efficiencies and improved results for yourself.

Jag Duggal is senior vice president of product management at Quantcast.  

On Twitter? Follow iMedia Connection at @iMediaTweet

Jag Duggal joined Quantcast in June 2011 and is charged with strategic positioning, development and management of the company’s targeted display advertising and audience measurement solutions. Prior, he spent five years in product management...

View full biography


to leave comments.