Blog

Intent Targeting vs. Contextual Targeting

Posted January 17th, 2014 by Rich Shea in contextual intent relevance   |   Comments.

Share:

The image above is an example of the problem with Contextual targeting with the type of unfortunate pairing of ad and content we all come across now and again. Look closely at the headline of the content behind the interstitial ad for “10 Best Cruise Ship Water Slides”, and you’ll see it’s about the passing of Gilligan’s Island star Russell Johnson.

How Contextual Gets Confused

There have been worse combinations, to be sure. But there’s little chance that a visitor will be interested in a cruise-related ad just because the subject of the underlying story is a star from a television show that featured as its set-up an ill-fated cruise.

The culprit here is so-called “Contextual” targeting. This ad technology scans the words or phrases in the content of the page, picks out the common ones, and then matches ads to pages based on a list of words or phrases associated with the ad. In this particular case it may have been the several instances of the word “island” on the page that triggered the match.

Really it’s a shame that this approach is called contextual, because in truth every fact about what is involved and led up to that ad impression is context, but that’s the terminology we’re stuck with.

Why Intent Is Better

In contrast, Yieldbot’s Intent-based targeting comparatively discounts the value of the content on the page itself. Instead, multiple Intent Signals, each derived from a unique “intent source”, are used to make a real-time determination of the appropriate ad for the specific pageview — if any.

To serve an ad, Yieldbot takes a handful of Intent Signals and brings them together to determine if any ads are appropriate to display for the pageview and which ads are most relevant. Sometimes it’s determined there actually is no ad relevant enough, and that’s fine too.

Yieldbot’s intent sources include data derived from external referrer links, the path of pages through the session, and key attributes of the page itself. Pages themselves have associations with intent based on sessions of past visitors and these are correlated with the signals for the current user for the pageview. An independent decision on matching is made every single pageview taking all of these factors into account.

Ultimately what makes intent better than contextual is that intent is about what this specific user is interested in right now. Pages themselves can be about many things, so at best contextual can parse that out and try to infer what the page might be “mostly about”. Intent brings into focus what actually interests this user, whether or not it’s one of the dominant concepts on the page.

Intent Enables Display Bought Like Search

In the Yieldbot Marketplace, Intent inventory is purchased in the same manner as Search (the grand-daddy of intent), through the use of keywords. Stemming from these dynamics, there are at least three reasons why Yieldbot would avoid the type of targeting faux pas illustrated above.

First, if contextual is about “what” (what is on the page) then intent is much more about “why” (why the user is visiting that page). The nature of intent is about the question: why are people (and this current person) visiting this particular page at this particular time? A frequently repeated word or phrase in the content of the page hardly factors into the equation. The keywords associated with the intent bringing users to that page is likely to be different and more relevant.

Second, negative keywords are a much more natural fit when thinking about intent. In this case it’s likely that a campaign about cruises, even if it had “island” as a keyword to match and that was coming through with an initially strong intent signal, would likely have negative keywords that would filter the ad from consideration (perhaps “die” or “food poisoning” for example).

Third, intent-based ad serving is inherently about relevance. As such, Yieldbot’s ad serving uses real-time updated performance data to aid in determining relevancy. If an irrelevant ad did get impressions for a short time, Yieldbot’s machine learning algorithms would determine that the ad was not relevant in that situation.

Moral of the Story

It’s all about relevance, and for relevance intent is king. For advertisers and publishers alike, serving a non-relevant ad impression is a wasted opportunity. As a result, advertisers in the Yieldbot Marketplace are seeing higher ROAS (return on ad spend) and publishers are getting higher CPMs.

— @shearic

Share:
comments powered by Disqus
More from Our Blog
Yieldbot's First Annual Super Bowl Intent Scorecard

Posted February 3rd, 2015 by Jonathan Mendez in intent, CTR

Yieldbot 2014 Review by the Numbers

Posted December 22nd, 2014 by Jonathan Mendez in

Rise of the Intelligent Publisher

Posted November 10th, 2014 by Jonathan Mendez in Media, CPC, Performance , Publishers, Data, Analytics, First Party, real-time

TF-IDF using flambo

Posted July 22nd, 2014 by Muslim Baig in Clojure, Data, flambo, Analytics

Marceline's Instruments

Posted June 25th, 2014 by Homer Strong in Clojure, Data, Storm, Analytics

View More

Yieldbot In the News
RTB’s Fatal Flaw: It’s too slow

From Digiday posted September 23rd, 2014 in Company News

Yieldbot Hands Publishers A New Way to Leverage Their First-Party Data

From Ad Exchanger posted September 23rd, 2014 in Company News

Yieldbot Raises $18 Million to Advance Search-Style Display Buying

From AdAge posted September 23rd, 2014 in Company News

Follow Us

Yieldbot In the News

RTB’s Fatal Flaw: It’s too slow

From Digiday posted September 23rd, 2014 in Company News

I have some bad news for real-time bidding. The Web is getting faster, and RTB is about to be left behind. Now, 120 milliseconds is becoming too long to make the necessary computations prior to page load that many of today’s systems have been built around.

Visit Site

Yieldbot Hands Publishers A New Way to Leverage Their First-Party Data

From Ad Exchanger posted September 23rd, 2014 in Company News

Yieldbot, whose technology looks at a user’s clickstream and search data in order to determine likeliness to buy, is extending its business to give publishers a new way to monetize their first-party data.

Visit Site

Yieldbot Raises $18 Million to Advance Search-Style Display Buying

From AdAge posted September 23rd, 2014 in Company News

Yieldbot, a New York based ad-tech company that lets advertisers buy display ads via search-style keywords, has raised a $18 million series B round of funding

Visit Site

Much Ado About Native Ads

From Digiday posted December 5th, 2013 in Company News

The most amazing thing about the Federal Trade Commission’s workshop about native advertising Wednesday morning is that it happened at all. As Yieldbot CEO Jonathan Mendez noted...

Visit Site

Pinterest Dominates Social Referrals, But Facebook Drives Higher Performance [Study]

From Marketing Land posted October 3rd, 2013 in Company News

Publishers in women’s programming verticals such as food and recipes, home and garden, style and health and wellness have found a deep, high volume source of referral traffic from Pinterest.

Visit Site

Pinterest Sends Your Site More Traffic, Study Says, but Maybe Not the Kind You Want

From Ad Age posted October 3rd, 2013 in Company News

Pinterest may have quickly arrived as a major source of traffic to many websites, but those visitors may click on the ads they see there less often than others.

Visit Site

From Our Blog

Yieldbot's First Annual Super Bowl Intent Scorecard

Posted February 3rd, 2015 by Jonathan Mendez in intent, CTR

Read More

Connect With Us

Where to Find Us

New York City

149 5th Ave.
Third Floor
New York, NY
10010

Boston

1 Clock Tower Place
Suite 330
Maynard, MA
01754

Portland

1033 SE Main St.
Suite #4
Portland, Oregon
97202