Running no-option pre-roll spots in front of online video content appears to be the best way to make money at this time. Research shows it’s the model best tolerated by viewers. 
The Seattle Times is doing this, but it making several awkward mistakes. I’m sure they are not alone – I just happened to catch these.
Mistake one is allowing full 30 second spots. The old 30 second spot designed for the days of appointment viewing over-the-air TV is dead. It’s too long, and often exceeds the length of the feature. If good news stories are now 90 seconds at the most, a 30 for a spot is two or three times as long as it should be.
The biggest mistake though is the gross emotional mismatches that are occurring between the spots, and the content of the video.
I was looking at their poorly shot raw video of the crime scene from the officer involved shootings last night in Eatonville. The grim crime scene is preceded by up-tempo, snarky and bouncy ad content promoting Google. It’s just terrible.
Television spots were never really “aware” of the nature of their surrounding content, but there was always a buffer – an anchor signaling that, “We’ll be back after this” or several other transitions and audio-visual reminders that indicated that the content and the ads were not related.
But with the pre-roll on the web, the spot is married to the single piece of content it precedes, and when the tone or emotional temperatures are off, it is truly awful.
This is a huge pitfall of this new advertising model that is going to have to be dealt with. My guess is that it could be done with metadata and an algorithm that better matches pre-rolled spots and the content. And maybe, just maybe, if something is particularly tragic – we could hold off on trying to monetize it.
In my view this has to be done, and fast – because some of the current results are embarrassing – just as bad as it is when the anchor smiles during a read about a murder.
Have you seen any examples of this?
