Wednesday, September 7, 2011

Lesson #88: The Basics of Online Display Ads

We have all seen graphic-based display advertising running on various websites.  Today's lesson is going to discuss: (i) when you buy these ads; (ii) the various placements; (iii) the various technologies; (iv) levels of targeting; (v) committed vs. remnant placements; (vi) approximate rates/expected clicks; and (vii) tracking, of these ads.

First of all, display ads are much less cost effective than other online marketing vehicles, like search engine marketing, email marketing or affiliate marketing.  So, I suggest not launching a display ad campaign until you have tapped into these other more cost effective vehicles first.  That said, display ads are designed more for use as a "branding" tactic, than an "marketing" tactic.  Which means they are used less for immediate transactions and ROI, and more for driving long term awareness and recognition for your brand.  And, most startups do not have the luxury of spending branding dollars, given the longer term payback.  So, plan accordingly, based on your available cash resources and desired timeline for customer engagement.

In terms of display ad placements, they come in numerous shapes and sizes and locations, almost too many to detail.  But, I will highlight a few of the more popular ones.  Most publishers that offer display ads offer them in three sizes: (i) the leaderboard at top of page (728 pixels wide x 90 pixels tall); (ii) the skyscraper on right side of page (120 wide x 600 tall); and (iii) the big box either in right column or interior of page (300 wide by 250 tall).  Other popular display ad placements include: (a) pop-ups in a new window over the page you are reading; (b) pop-unders in a new window under the page you are reading; (c) interstitials inserted in-between two-pages you are reading; (d) within an emailing; and (e) within or around a video player.  These latter examples are bigger, more productive placements, and typically come at premium rates than the first examples above, which we will detail further below.

The creative technologies you can deploy within these ads can be as simple as a static image, to moving parts (with technologies like Eyeblaster), to expandable size when rolled over with the user's mouse (with technologies like Pointroll), to data-entry forms for sign-ups, to moving video.  There are also pullback technologies you can deploy, that drops a cookie on the user's PC when they first visit your site, and then pushes them ads on third party sites after they leave your site, to try and get them back. As you can imagine, the more complex the creative is to build, the more expensive it is to run.  As an example, producing a video is much more labor intensive than creating a static image ad.  And, using technologies like Eyeblaster or Pointroll, comes at the additional cost of $3-$5 cost per thousand impressions (CPM) to serve ads using their technologies.

The great thing about online display ads is you have many ways to target the placements of your ads.  You can either run ads run-of-site (ROS), on any page of a publisher's website.  Or, preferably, you can target your ads based on: (i) demographics of the user; (ii) subject matter of the content (or behavior of the user); or (iii) time of day/week, to name a few.  But, although the performance of deep targeting is materially better than ROS ads, publishers typically charge a premium for each level of targeting you desire.  For example, a ROS ad for "general travel" may cost you $2-3CPM and a deep targeted ad for "rich adventure travelers looking at Kenya content" could cost you $10-$20CPM, depending on how many levels of targeting you desire (e.g., rich, adventure, Kenya in this example).

There are typically two ways to buy display ads.  One is on a "committed" basis, which means your order secures a guaranteed placment in the time period you desire.  The other is on a "remnant" basis, which means your order will be run only if there is unsold space available in that period (which means it is not guaranteed to run if there are committed advertisers willing to buy the space).  Committed placements come at a much higher CPM than a remnant placement.  But, when you can get it, remnant ads are definitely the way to go for a startup, given the material cost savings.  And, publishers should be happy to consider remnant buys, as getting a little bit of revenues at a much-reduced rate, is better than letting the space go unsold for zero revenues.

Typical CPM rates for display ads come in a very wide range.  They can be $0.50CPM for a static graphic ad on a remnant ROS basis, to $10CPM for a committed/targeted basis, to $35CPM for a video ad, to $50CPM for a targeted email inclusion.  Of course, rates are typically tied to expected click performance of these placements, where clicks can be 0.05% for that cheap remnant ROS ad, all they way up to 5.0% for a targeted email inclusion.  And, don't forget, CPM's and clicks also rise for using various technologies or targeting your placements, as we discussed above.  As an example, a static ad may get a 0.2% click rate and a moving/expandable ad may get a 0.6% click rate.  Or, a ROS ad may get a 0.2% click rate, and a deep targeted ad could get a 0.6% click rate.  So, make sure you ask about expected clicks from each placement, to justify any higher CPM's.  And, where you can, ask for CPC (click), CPL (lead) or CPA (acquisition) advertising opportunities, which is always better than CPM (impression) based placements, since the user has made a known action on your site.

And, finally, the best thing about online display ads is the ease of tracking the results therefrom.  You should be using unique tracking codes for each specific placement you buy, to track the effectiveness of which sites/placements are performing better or worse than others (from a small test campaign).  And, then optimize the bulk of your remaining budget/campaign into the best performance vehicles from the initial test.  And, where you can, tracking should be implemented at the click, lead and transaction level, to determine which clicks/vehicles do a better job of leading to customer engagement and conversion.

For future posts, please follow me at:

No comments: