1/ It’s prediction time on contextual advertising. The burning question: “What happens to publisher ad inventory when 3rd party data gets taken away?” https://twitter.com/garjoh_canuck/status/1166733407776399360?s=20

2/ Everyone is applying a loss of 3rd party cookies to the publisher content and media formats of TODAY. Contemporary publisher content has been optimized FOR 3rd party cookies—it entirely stands to reason publishers would be adversely affected by cookies’ disappearance.

3/ Sure, publishers can and will make up some of their losses related to 3rd party cookies via subscription revenue, but I think we all know that most publishers won’t attract paying subscribers in sufficient numbers. It’s unrealistic. https://twitter.com/WebPublisherPro/status/1318601876070404099?s=20

4/ There is also the “first-party pub data will replace 3P data” story. Yes, 1P pub data is valuable, but very, very few publishers have such data at sufficient scale to really monetize it efficiently and make it attractive to advertisers when compared against the Walled Gardens.

5/ Now...I am not saying publishers shouldn’t focus on 1P data. There are lots of opportunities there. But given that 1P data is expensive to activate and is mostly valuable only at great scale (e.g., NYT scale), on it own it’s unlikely to “make publishers whole” again.

6/ All of these notions assume that publisher content formats stay static in the face of 3P cookie loss:

1) CPMs will plummet when 3P cookies go away!

2) Paying subscribers will save us!

3) 1P data will save us!

Again: WHY ARE WE ASSUMING CONTENT STAYS STATIC???

7/ A lot of shade is thrown at contextual ad inventory. Sure, contemporary contextual mostly sucks, because much of the time a “contextual” display or video placement is just a regular placement stripped of the ability to apply 3P data to it at auction time.

8/ Or it’s those same 3P-data-optimized placements being sold via direct or site buys.

Do you see what I’m getting at where everyone is expecting 3P-data-optimized placements to somehow effectively do double-duty as contextual placements?

WHY??? Why would we expect this?

9/ Here’s a truly contextual ad placement. Just grabbed it from Google Maps. It takes meaningful contextual account of me, the viewer. Google didn’t shoehorn IAB units into Maps; it made sure the content, context, and ad were all symbiotic.

10/ Publishers can’t all pivot and survive by “being more like Google.” That’s a completely unrealistic expectation. There is a reason Google is so dominant—that level of consumer intent is really hard to capture.

11/ But publishers CAN evolve their content to make sure it is more contextually compatible.

12/ Content will NEED to map to monetizable segments...TBH just audit the behavioral segments your advertisers are currently targeting when they buy your inventory using 3P cookie data. Not sure how feasible this is, but it seems like a logical base to cover.

13/ You could get down to the level of using navigation and page design to throw off streaming contextual signals in real time that can be passed into bid requests.

14/ Most sites aren’t designed that way currently. Why? Because historically it’s been much easier to have fairly static pages and ad placements that can be easily monetized at auction with 3P data overlays. But times are changing; maybe pages and nav should change too.

15/ Video and audio are two other exciting areas. Very much still in the dark ages, but will come into their own as contextual advertising is forced to evolve. https://www.google.com/amp/s/thehustle.co/09162020/amp/

16/ Do most publishers offer audio versions of all their content? No. Why? Easier to do screen-only, throw some IAB placements on the page, and let 3P data do its trick. But what if I’m out on a walk? Publishers are sitting on a gold mine of contextual audio data.

17/ Going back to navigation. What am I potentially telling a publisher as I browse a site? Probably a lot!! Does the site not throw off enough navigational data? Well then completely reimagine the navigation until you get the contextual data you need!!

18/ For example, maybe big content-heavy pages are bad for contextual. You have a bunch of ad placements that all need to be filled simultaneously and as a pub you get one page load to “guess right” as to what to serve.

19/ Why are so many pages like that (content-heavy)? BECAUSE 3P DATA MADE IT FEASIBLE, not because it’s an inherently superior approach. And yes, when you apply contextual inventory to that content, it kind of sucks.

20/ Maybe break up the content. Keep visitors DOING STUFF (browsing, tapping, moving cursor, choosing from menus, etc) so that you get more data about sessions. Give everyone a miniature “rabbit hole” to go down. But a rabbit hole that’s relevant to your site, of course.

21/ To be clear, I am not advocating that publishers just replace all content with Taboola-inspired click-baity chum.

22/ For example, if you publish a feature, break it up into “chapters” with respective ad placements, or write the feature from a few different perspectives or interest areas that can be chosen from, allowing the reader to self-identify via their own browsing behavior.

23/ Anyway, I’m about done with this rant. Just wanted to jot down thoughts on how publisher CONTENT can evolve as a way to “rebalance” the equation in a post-cookie world. It just annoys me that we’re trying to apply solves to publishers that are assumed to be frozen in time.