The “living room tax” is such a good way to frame this. CTV targeting can be incredibly precise, but precision doesn’t always mean the impression is valuable.
One thing I’d add is the quality of the traffic behind those impressions. Even with the right account, context and timing, it still matters whether you’re reaching a genuine viewer and getting a valid impression. CTV really needs both sides of the equation: smarter targeting and better traffic quality.
There’s no doubt that TV done right can be incredibly powerful. B2B needs some consideration which B2C doesn’t because it’s bringing work into the house and into lean back private time. Doesn’t mean it can’t work but some guardrails makes sense.
Good points on traffic quality too. This is why I like contextual YouTube (where we know the exact video context) delivered on the big screen. Best of both worlds and more B2B inclined
It backfires in my opinion when we buy it as we are buying any other digital channel- squish the precision on every single data point for targeting while you are leaving 80% of good inventory on the table.
Just because it is programmatic doesn’t mean it has to have layer over layer over data targeted to a tiny fraction of your actual ICP.
We can learn a bit from actual TV buying here. This doesn’t mean you go to a shopping list of buyers, you still target but don’t lazer focus on data, especially if it not built on your first party!
Yeah the reality behind all the noise is that programmatic tv is still new. Upwards of 70% of ad requests are empty of the metadata that enables precision targeting - which validates your point on not layering data up because the volumes of available inventory become really very small. I see good applicability in tv sponsorship extension, taking for example F1. B2B brands anchor to it but there’s wider F1 content out there across the various “tv” modalities including some very premium and expert built YouTube channels.
I do think we have to be very careful to not take our signals and blast them widely in TV, especially in out of hours times of the week. Any value in that precision is lost if the user is left p***ed off!
I agree with you. It is relatively new. Now what is also happening is direct buys because inventory quality and well DVF or IAS being very subpar products unfortunately as we all have known for better part of the last decade.
Open Path relatively seems better, that is also depending on case by case basis.
The best alternative to all of this is using a custom model on your first party that you own with Chalice for instance so you get the quality, the proper targeting (so no snoozing or else) with proper ability to measure causality.
I would take it further than CTV and say work with Chalice- containerization. Viola:)
Yes we’ve been very bullish on containerization. We migrated our account graph into containers to get cleaner signal on the “right side” of murky DSP auctions. We’re exploring DSP containerization too to save signal leakage. Adam and the Chalice team are doing some excellent work too.
The “living room tax” is such a good way to frame this. CTV targeting can be incredibly precise, but precision doesn’t always mean the impression is valuable.
One thing I’d add is the quality of the traffic behind those impressions. Even with the right account, context and timing, it still matters whether you’re reaching a genuine viewer and getting a valid impression. CTV really needs both sides of the equation: smarter targeting and better traffic quality.
Thanks for stopping by and I’m glad it resonates.
There’s no doubt that TV done right can be incredibly powerful. B2B needs some consideration which B2C doesn’t because it’s bringing work into the house and into lean back private time. Doesn’t mean it can’t work but some guardrails makes sense.
Good points on traffic quality too. This is why I like contextual YouTube (where we know the exact video context) delivered on the big screen. Best of both worlds and more B2B inclined
It backfires in my opinion when we buy it as we are buying any other digital channel- squish the precision on every single data point for targeting while you are leaving 80% of good inventory on the table.
Just because it is programmatic doesn’t mean it has to have layer over layer over data targeted to a tiny fraction of your actual ICP.
We can learn a bit from actual TV buying here. This doesn’t mean you go to a shopping list of buyers, you still target but don’t lazer focus on data, especially if it not built on your first party!
Yeah the reality behind all the noise is that programmatic tv is still new. Upwards of 70% of ad requests are empty of the metadata that enables precision targeting - which validates your point on not layering data up because the volumes of available inventory become really very small. I see good applicability in tv sponsorship extension, taking for example F1. B2B brands anchor to it but there’s wider F1 content out there across the various “tv” modalities including some very premium and expert built YouTube channels.
I do think we have to be very careful to not take our signals and blast them widely in TV, especially in out of hours times of the week. Any value in that precision is lost if the user is left p***ed off!
I agree with you. It is relatively new. Now what is also happening is direct buys because inventory quality and well DVF or IAS being very subpar products unfortunately as we all have known for better part of the last decade.
Open Path relatively seems better, that is also depending on case by case basis.
The best alternative to all of this is using a custom model on your first party that you own with Chalice for instance so you get the quality, the proper targeting (so no snoozing or else) with proper ability to measure causality.
I would take it further than CTV and say work with Chalice- containerization. Viola:)
Yes we’ve been very bullish on containerization. We migrated our account graph into containers to get cleaner signal on the “right side” of murky DSP auctions. We’re exploring DSP containerization too to save signal leakage. Adam and the Chalice team are doing some excellent work too.
Adam used to be my boss at Razorfish and I worked with Ali in the same search team…. Oh god, 2 decades… I am old:))