Thursday, May 23, 2013

Capturing Value from High CPM, Low Fill

After publishers start to surpass a certain threshold of traffic, many start to get a lot of inbound requests from verticalized ad networks who are able to offer significantly higher CPMs than their run of site network. It's important to recognize that the CPM isn't the only factor in making a decision to run these tags.

Vertical Networks: An Overview

Vertical Networks came about as a way for demand sources to band together around an industry vertical and gain access to similar inventory sources. The idea here being that buyers wanted some sort of scale in the form of multiple different properties centered around a vertical, but also wanted to reach a somewhat pseudo-targeted audience by virtue of the content on that inventory, without the hassle of entering into direct relationships with those publishers. As the industry has evolved, many vertical networks are rebranding themselves as programmatic exchanges to ensure they are not losing out to RTB demand currently dominating the market. 

Publishers Have Power

In a fairly rare instance where publishers have negotiating power at the table, it's always best to understand the fill rate that a vertical network is able to provide. Generally speaking, fill is going to be low (anywhere from 10-30%). The CPMs may be high, but a low fill will ultimately cost publishers who decide to pass all their traffic to these third party demand sources.

The Optimal Setup

Recognizing that there is a pricing histogram whereby impressions are valued differently along a curve, we can see that these networks should be trafficked so that publishers can capture the high CPMs offered yet monetize the remaining inventory via an exchange. In this vein we recommend publishers to set a fairly aggressive frequency cap to ensure they are able to capture incremental value and ultimately add to their bottom line. 

Tuesday, May 21, 2013

A/B Testing via Optimizely

Recently a lot of discussion for publishers has centered around how best to serve the user while also driving revenue growth from advertisers. In a multi-screen, mult-format world, how best should the sell-side adapt to changes?

A/B testing has always been the process by which savvy publishers make changes to their site which enable publishers to make data-drive decisions that will ultimately support any metric or goal that they wish to attain. Everything from user interaction, revenue per visitor, engagement and so on can be measured by a properly structured A/B testing framework. For publishers, having a trusted, easy to use A/B testing framework will ultimately pay off many dividends now and in the future.

What is Optimizely?

Optimizely was conceived as a platform to easily allow publishers to run multi-variate tests on their users to see how users interact with their site. For example, publishers can setup a test as simple as whether or not to use black or blue for a link color, or as complex as using a left navigation column layout versus a horizontal column layout, all while monitoring any number of user metrics, stats and KPIs. Many major publishers are already using A/B testing in some form or another, but in many cases these frameworks took a massive amount of time and engineering resources to get up and running, and many publishers simply could not make the investment.

Optimizely democratized the A/B testing movement and opened it up to the masses; no longer do publishers need a large engineering team to build out separate code bases and tests. Optimizely easily allows you to create different variations of any page layout, and push forward a day parting schedule of your choosing. Users can collaborate on the same experiment, test out various changes, and implement them to a live audience in a matter of minutes. The only technical lifting that needs to be done is the initial upload of a single line of JavaScript just inside the opening <head> tag. After this, Optimizely then renders a page layover for the user and tracks metrics accordingly.

To get started, head over to Optimizely and take a look at their offering.

Friday, March 1, 2013

Account Dashboards as a Monetization Opportunity

Many publishers have a login feature (either using social login like Facebook or Twitter or using their own homegrown system). For publishers, these login systems can be a great way to gather information on their audience and perhaps provide them with better targeting. That being said, these account dashboards also provide a unique opportunity to monetize a user given the intent of the user when they log in.

A User's Intent Drives Clicks

It can be reasonably assumed that when users are logging into an account on your site, they are doing so for a purpose: checking alerts, updating their payment information, setting up a new billing address, etc. In fact, users are really logging in to perform a very specific action. This is really the underlying reason why they could be monetized so well with a properly placed advertisement.

Even if the user hasn't already completed the task they are doing, and it isn't that important to the user at that time, the advertisement provides a compelling opportunity to monetize the user upon exit. If the user was already going to navigate away from your account because either a) their task is completed or b) they felt compelled to leave and take care of it later, it would be best to monetize both options using an advertisement.

Placement, Placement, Placement

Monetizing your users on an account dashboard will provide a great way to properly monetizes users once they are ready to exit a page. The placement can really define whether or not the ad performs well or under performs. Central to this idea is that we are not trying to bombard the user with advertisements; rather, we are trying to place the advertisement in such a way that they are able to discern the message without really harming their user experience flow while completing tasks within the account dashboard environment. When executed properly, publishers can more effectively monetize this audience of users, and grow revenue as a result.