Same ad, three placements, different results.

That's why online ad targeting matters.

In our example case, the ad was for a major skin-care line. As part of a test, conducted by OTX Research, the ad ran in a site that was non-endemic, meaning a non-health and beauty location; an endemic site with lots of content about skin; and on an endemic page within a neutral site (such as an article about skincare on a general-interest newspaper website).

The last two placements showed about a 19% lift in brand recall over the former, proving that when it comes to online ads, contextual targeting can have an effect.

"There has been no definitive study done on the effect of contextual targeting -- it's mostly been intuitively taken for granted that it works," said David Brandt, managing director-marketing insights at OTX, a global consumer research and consulting company. "But we haven't seen a downside: The effect is either a lift or neutral. It's nice when life reacts the way you expect it to."

Contextual targeting, one of the web's most popular ways to match an ad, was popularized by Google. Say you're reading a story that describes Baja, Mexico, off-road racing; the page might display sports cars, products for race-car enthusiasts or even offer trips to the Baja peninsula.

Google AdSense Director Kim Malone Scott said that "context and scale have allowed us to re-create the mass market."

Contextual shouldn't, however, be confused with search-based targeting. "Search-based advertisement is about demand fulfillment, and contextual advertisement is about demand creation," said Jay Sears, senior VP-strategic products for ContextWeb.

Deeper targeting
Site targeting is a kissing cousin of contextual targeting; it lets advertisers place their buys on websites that are related to the product being marketed. Ad networks usually aggregate publisher websites into categories and offer these categories as bulk packages to be sold. Site targeting can also aim at specific pages within publisher sites. It reaches people where they congregate around shared interests and in many ways represents the overwhelming bulk of site inventory.

"For niche products, affinity is very effective," said CheckM8 Chief Operating Officer Oren Netzer.

Targeting context is just one way online marketers are trying more efficiently to match ads to a relevant audience. Increasingly, marketers are also targeting ads to people. According to Marketing Sherpa data, 40.5% of marketers said contextual targeting delivers good return on investment; behavioral targeting was not far behind at 36.7%.

The granddaddy of that kind of targeting is demographic, which marketers have been using for years offline, based on the idea that consumers are defined by their personal characteristics, including age, gender, income, ethnicity and other variables. But collecting the appropriate data online can be challenging.

"Very few sites are doing demographic," Mr. Netzer said. "You need user details for that, and that requires asking for it directly from them" -- through registration, for example.

Pitfalls
Of course, sometimes a person's interests can be discovered more implicitly -- by using their online patterns to discern interests and behavior. Behavioral targeting aggregates data retrieved from cookies and data so that trends appear. This can be based on past behavior -- spending lots of time at parenting sites, for example, probably indicates a person has a child -- and predictive modeling can be used to display ads that people are most likely to find interesting. Behavioral targeting, however, has elicited privacy concerns, especially as new players aim to tap into behaviors at the internet service provider level. Don't target just for the sake of targeting, advises Michael Cassidy, CEO of Undertone Networks. "You can find the person that drinks Snapple with red shoelaces, but you'd be reaching ten people at best. ... You're better off starting broad and develop into a narrower niche," he said.

Retargeting (or remarketing, as it's called in the offline world) is often classified as part of behavioral targeting. In fact, some networks that claim to offer behavioral are really offering retargeting, which aims to locate consumers who dropped off midway on the path to a purchase and serve them a new ad in the hopes they will complete the purchase.

Purchase-based targeting is based on information about purchases made on publisher sites; the nature of the purchase then informs further offers.

Offline, this dynamic has been developed by companies such as Catalina Marketing, which tracks consumer purchase history using loyalty cards and then uses that information in real time to print coupons with personalized offers along with the customer's receipt.

"Do you want to target people who want to be healthy or those who buy $250 of health products a year?" noted Catalina Senior VP Todd Morris. "If a person buys granola bars, we'll print a coupon with an ad for a brand with twice the protein."

Targeting at a Glance

Demographic: Traditional targeting based on data such as age, gender, income and ethnicity.

Geotargeting: Targets a consumer in a certain geographic area using location data mined from the ISP or IP address. A powerful tool for local business objectives and marketers, it allows the display of local product inventories to customers in the case of new rich-media ads.

Behavioral: Tracks the actions of thousands of users as they surf the web and aggregates them for trends. These patterns become the basis for targeting and can include purchase history. One example: A visit to Edmunds.com can be the basis for serving an auto ad even after a consumer moves onto a non-auto-related site.

Contextual: Matches the advertising to the content being consumed (whether a text, audio or video file). For example, an article about the outdoors brings up a camping ad.

Site-targeting: Similar to contextual, this tactic matches the ad to the theme or genre of a publisher. A high-definition TV marketer may choose to advertise on sites dedicated to consumer electronics.

Daypart: Just like offline, daypart focuses on people's work/life schedules. One example: targeting ads for Egg McMuffins from 8-10 a.m. Daypart targeting also works well for impending product releases and roadblock campaigns.

Purchase-based: Tracks the purchase history of users to establish trends, much like behavioral. People who bought one brand's shoes might be interested in more of the same or of another brand.

Retargeting: Aims to locate consumers who dropped off midway through the path to a purchase and serve them a new ad in the hopes they will complete the purchase. Called remarketing in the offline world and sometimes classified as part of behavioral targeting online.