Generally when you buy advertising on digital properties you are either buying on
– context (ie News, Finance, Sport)
– performance (Cost per whatever)
– Targeting (day, time, demographic, behaviour sub category, browser, OS etc)
– or a combination of all 3 …
Sometimes it seems to me that there is a belief that within a category all ad impressions are created equal. Thus, you pay $x cpm to appear in ‘Local News’ regardless of whether your ad appears on an index page, within a story, below the fold, above the fold, on a slideshow, on editorial or on feed provided content.
This seems weird to me. It’s obvious that there’s massive differences in the quality of ad space within this one environment. All impressions aren’t created equal.
A few months back I conducted a workshop with a publisher and we talked about this very issue. They generally sell their site as a run of type option and also have the potential to target based on user registration data (as registration is required for all users).
One thing we talked about was the idea of packaging up inventory differently to the current norm. Looking at how users navigated through the site and passing on these learnings to advertisers.
For this site the pattern was similar to the below
Homepage entry point – User enters name and password. Spends around 10 seconds on this page
User homepage – dashboard for all key information and gateway to site. User would spend around 30-40 seconds here
Page X – User enters in key information. Spends around 3-4 mins on this weekly
Page Y – User can check other information and compare information. Spends around 2-3 mins weekly
Confirm page – User is in and out – maybe 5-10 seconds
Logout – User spends 2-5 seconds here.
The issue was, inventory was being charged at the same dollar value regardless of the dwell time of the consumer. I found this weird – the opportunity of 2-3 seconds is way less valuable than the opportunity of 2-3 minutes.
What I suggested was tiering the content and being clear to market the rationale behind these tiers and the pricing.
This concept I believe could be looked at across all publishers as many would have similar traits.
– Homepage – user is transactional, looking for content with purpose.
– Within articles – user is more engaged, reading content and spending more time on the page
– Slideshows. User is churning through a pageview every 1-2 seconds … it’s all about immediate gratification
– Features versus feeds. One could reasonably assume someone reading a feature is more engaged than someone reading a 100 word AAP feed
I think with tiering you could address the issue of CPM paying advertisers rubbing shoulders with performance/cheap CPC advertisers. The other day I saw an ad on Vogue.com.au (womens lifestyle) for Ashley and Martin (Mens Hairloss). How ridiculous? I was informed it was a performance ad but still … who are the people thinking it’s a good idea to have high end womens lifestyle brands like Chanel etc running next to Yeah Yeah ads?
With tiering it would add transperency and advertisers would know upfront where they’re running (ie in articles, in index pages, on slideshows) and whether they’ll be running next to performance ads for hair loss and Optionetics.
My belief is this would ultimately include yield as advertisers would be willing to pay for more control over placement. Right now, the belief of many is advertisers will be ‘gamed’ by the publisher and that pageviews are being boosted by cross network linking tricks, slideshows and additional ad placements.
If a publisher came to me and said – “Ben, I can guarantee you will run only within article text and you won’t run next to any performance advertisers” I would be more than happy to look at paying a fair premium and I’m sure others would too.