Re your note about segmenting and analysing by differing levels of user engagement:

This is something I have hands-on experience with. We segmented by the source of the user sign-up (ad campaign/promotion partnership/traffic referral etc), looked at the engagement of each segment over the following month, and having identifed our most valuable source of traffic (not by volume, but by actual VALUE – persistent engagement with content as reflected by advertising revenue) we could focus accordingly.

Some more datamining thoughts are at: http://platform.idiomag.com/2009/06/the-goldmine-of-reader-data/

Think about it, you just determined what pieces of content, types of content and/or editors (or whatever else dimensions you have per content piece in your CMS) draw the most engagement from your users. And so now your 2 questions are going to be: well, where are those most engaged users coming from and how can I get more of them! This is given you already proved that Engagement = Value for Advertisers. To prove Engagement = Value for Advertisers you create 2 visitor profiles: profile_1 that are Engaged, and profile_2 – disengaged and compare them against your bottom-line metrics site metrics and advertiser metrics…

I’m working on a presentation for the eMetrics summit in DC, Oct 19-22, specifically on this topic. Do you want to weigh in on it? I can send you a raw draft. Once I have it that is

But the issue with new ways of measuring performance is that you need a substantial amount of advertisers to buy into it (and therefore spend time and money deviating from their normal ad buy process). So it would probably be more effective to extend your measurement of engagement all the way to reader action (PPC, CPA, CPE etc) – which is something that advertisers do pay for.

Measuring Engagement is an extremely difficult task, but there are ways to come up with Proxies based on each Publisher’s specifics and requirements. If you can prove how Engagement affects visitor behavior, than you have a model that can allow you sell Audiences based on their Engagement Index as well sell Content based it’s Engagement Index.

However, experience offers ready opportunity to create uniqueness and thus monetarisable value in the perception of the audience. As humans, we do this all the time – on one hand dynamically adapting how we talk to one another to get our point across or persuade and on the other “valuing” each experience in terms of pleasure of the interaction and the quantity, quality and reliability of what was gained. The challenge is to create a similar experience through the published medium. More thought required.

True/Slant is one model that’s certainly interesting – and brings together elements of the Le Monde model, branded content and custom publishing. In short, advertisers will be able to maintain content areas within the site – essentially creating content in the same fashion as the site’s paid journalistic contributors.

Offline, I’ve recently written about Monocle magazine as an example of an innovative model that blends many of the ideas above. They’re leveraging the print magazine to generate sales of physical products developed in partnership with boutique firms – sold via the magazine, the mebsite and 3 physical shops.

It’s doubtful that any one business model will provide the ’silver bullet’, making the case for blended models all the more relevant.

John – agreed. Trouble is its so easy to write link-bait content. Some publishers seem to be focusing on making a quick buck by writing content based on Google Trends, or endless Top-10 lists, which certainly does bring an immediate kick, but probably hurts the brand long-term…