I came across an interesting post on Mashable last week, written by Jitendra Gupta, the CEO of internet startup SezWho.

In it, he focuses on social media advertising but also addresses the broader issue of CPM-based advertising. It's this part that I feel is worth discussing.

In his post, Gupta argues that the reason a CPM-based advertising model doesn't work on social media properties, such as social networks, is that this model tries "to target new media users using old media technology and thinking."

He goes on:

"The CPM model asks, 'What is the content on the page?' But social media conversations aren’t ON any one page anymore. They jump from one site to another, they take place on a variety of platforms and media types, across blogs and Twitter and FriendFeed and Facebook and MySpace."

Gupta's solution to this conundrum - universal profiles that follow users wherever they go. A convenient solution since his company has built "a universal profile service for the social web that improves community engagement and enables content discovery to be added to blogs, forums, message boards and other social sites."

I, of course, would argue that a universal profile service is about as viable as putting a man on the sun, but let's take a step back and look at the big picture.

Most online publishers and ad buyers know that the CPM advertising model doesn't work. The average publisher can't command premium rates (especially when stuck with ad networks) and the average ad buyer doesn't see an ROI. I touched on this in a post early in January.

Yet ad buyers will still spend on CPM deals in large part because CPM is a known quantity. Brand advertisers understand it and are comfortable with it. After all, they've been paying for "impressions" in some form or another since well before the internet even existed.

In today's economic environment, of course, it's no surprise that a lot of the CPM ad spend has already started fleeing to quality. Search advertising, obviously, is a major beneficiary because of the perceived accountability and control.

The challenge for individuals like Gupta who are trying to develop and market "innovative" new online advertising models is that they're really not addressing the problem.

Gupta's solution is essentially targeted advertising wrapped around "user engagement":

"Context-based discovery and social search play a role in bringing relevant content to receptive audiences. Interest-based user engagement — based on user context, not content -– tracks interest across various conversations and among the plethora of social media sites."

"Knowing – based on her participation across social sites – that a social media participant is a New England Patriots fan who vacations frequently in Hawaii and loves water sports and high-tech gadgets is invaluable to creating a tailored interaction between that participant and companies with offers that are relevant to her."

Behavioral targeting, anyone?

As I've reported on before, there's some evidence that behavioral targeting isn't all that it's cracked up to be. And this makes sense. The reality is that delivering the right ad is easy but delivering it at the right time can be incredibly difficult, especially on websites where users are interacting with each other. One of the reasons search advertising is so effective is that the timing is right.

Gupta's model does nothing to address some of the fundamental concerns ad buyers have about measurability. As one commenter pointed out:

"Given that CPM on a clickthrough and impression basis has been falling off a cliff in terms of rates and effectiveness, how exactly will advertising look? What sort of packages and pricing will be offered to advertisers? How will the effectiveness of this contextual advertising be measured and benchmarked?"

"The underlying problem is that the inherent measurability of action and activity on the internet has led to undervaluing the advertising that runs across it because value created but not acted upon (brand exposure, etc) is not really measurable."

All valid points.

Specifics aside, I think a very real problem when it comes to innovation in online advertising is that people are trying to be too clever.

Individuals like Gupta, who it's worth pointing out has a technology background and not an advertising/media background, seem to be looking at the challenges in the online advertising market from theoretical and technological perspectives.

Gupta claims that with his model, "advertisers can now understand and appeal to an individual’s online and offline interests based on the fullness of their participation across social media sites."

That, of course, sounds great in theory but how is this sold to an advertisers? How does an ad buyer quantify what he's going to get? How is everything packaged into a deal that's compatible with the economics of the ad buyer's business? How does the seller make the ad buyer feel "comfortable" with the deal?

The bottom line is that few people are considering the needs of advertisers. In general, I see a lot of discussions of this nature - "ad model x doesn't work on website y so how can we create z to fix website y's problems?"

What this really means is that individuals are trying to create new demand with a new model. Instead, they should be trying to create a new model that will meet existing demand.

I've considered the social media monetization issue and I've generally come to the conclusion that there's far less money to be made here than many previously projected because the value of social media properties as scalable advertising platforms is limited.

That said, social media hardly presents the only online advertising challenge.To those trying to solve any of the many challenges in the online advertising market, a word of advice - ad buyers may not be satisfied with CPM but your sales pitch has to be more than "CPM advertising sucks."

Trying to outsmart the money flow in any market is a losing battle and if you look at where money is going in online advertising today, there are certainly some places it doesn't make sense to be.

Drama 2.0

Published 28 October, 2008 by Drama 2.0

237 more posts from this author

You might be interested in

Comments (5)


Paul C

Also what everyone seems to forget is that the vast majority of advertising is for everyday products like insurance and its all very well monitoring the fascinating interests of consumers but 99% of it is irrelevant to most brands.

almost 10 years ago


Picky Boot

CPM is one of the best online advertising program. It is not that all the advertisers doesn't gain anything from CPM ads. There are so many advertisers who make a good business with this advertising network. Few people actually do not know to make CPM advertisements to the extent, which would give them the business. Ya the people are trying to be clever in online advertising that is real problem.

over 9 years ago


David Ogilvy

CPM only works for large networks like Google or DoubleClick ... 

what if u r getting 10,000 looks per day? thats better than an OUTDOOR board! (which cost THOUSANDS per month!)
best to sell ure ads door-to-door!  there's no value for the independent guy.

over 9 years ago



CPM works good for most of the sites very few they go bad. If they site is not having good traffic then CPM really goes bad. Since rate is for thousand impression the site should have a good number of traffic everyday. If the site is having good traffic then I am sure that one make a great ROI.

over 9 years ago



Dude you are an idiot. CPM based ad buys can be some of the most lucrative campaigns you could ever imagine. You don't know shit if you think CPM buys don't work.

over 9 years ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.