Excellent write up Omar, and great as I was not able to attend! All included is insightful yet not surprising. I just feel though that we’re gonna be reading the same thing in years time. Tv measurement horses have bolted in their different directions now and not sure if incentives and legacies can ever truly align to what great effective tv measurement looks like. Device is touted but traditionals won’t let go of what is serving them in their dwindling. It’s all misaligned incentives. Thanks again (and look forward to reading the not-too-similar-I-hope next year)!
The broadcaster language is certainly changing around devices and, as Sameer politely reminded me (see below), Lantern is supposed to be an answer to this. In principle, that shouldn't be a problem. Whether you're sampling real-world audiences, or counting actual device impressions, I don't see why advertisers shouldn't expect a) to know how many people *really* saw their ad; b) what impact their adspend has on sales. Advertisers need to be more aggressive in demanding this information, not accepting what media owners and platforms say as gospel because it makes their lives easier.
Comment from Sameer on LinkedIn (complimentary of this piece but I've edited for brevity):
"The future is about fair allocation of money relative to the value created for advertisers. And on that we have nothing to fear."
"I slightly wish you’d noted that this is what Lantern is all about. We may be feeling pressure, sure, and at the same I’d probably say the response is more purposeful than panicked. Pilot is done. Beta is revving up. Transparency is absolutely going to be part of it... but it will be a different take on transparency than you need when doing counting, because incremental lift is more complicated.
"We need to skate to where the puck will be - in this instance the platforms are heading to near-term incremental outcomes, and so are we."
In previous posts I said publishers need to be transparent about how they use AI, given that trust will become even more important in the age of AI slop and misinformation.
So, as a self-publisher, let me tell you exactly how I've used AI for this piece.
1. I used Otter.AI to record what was said at this conference. It produces a transcript which is of variable quality in an echoey event chamber, so I had to double-check the quotes I used in the article by listening back. But Otter is great at letting you follow along.
2. I pasted each transcript into ChatGPT to give me a summary of what was said. This is sort-of helpful to make sure you didn't miss anything in my notes.
**At this point, it's important to stress that during the talks, I was taking notes, but I wasn't doing the analogue strenography we're taught by the NCTJ i.e. I wrote down key points/phrases by speakers on one half of a page, but then sometimes my own reaction/thoughts/observation on the other half. So the commentary is forming in real-time, not just after the event. FWIW I use an iPad, Apple Pencil and the Notability app, which is slower than typing but I like to doodle and draw arrows.
3. The day after the event I spent the best part of an hour just freewriting what I thought about what I listened to. Freewriting is an important part of my practice and (touch wood) it has basically eliminated writer's block for me...
4. ...which is dangerous because you can produce a lot of shit writing when you get rid of your filters. A good writer is a good editor. I pasted my freeform writing into ChatGPT and asked it to suggest a structure for an essay. I didn't like the bitty format it suggested (even though it was likely more optimal for the reader who just wants to scroll and not actually read anything), but I did keep the 3 takeaways it suggested at the end. But broadly as narrative flow it didn't deviate from my freewrite much anyway.
**At this point it's important to say that the pictures are not just decorative; they are supposed to help tell the story. This is the other advantage of not being a stenographer in the room; I am able to think about taking pictures strategically. Pictures are incredibly important. I use an iPhone 17 Pro which, despite having a shitty aluminium casing which literally scratched as I unboxed it, has an INCREDIBLE camera for a smartphone (look at the panel closeups; I was 10m away in a dark room!) - tech also powered by AI.
5. Then comes the fun part: turning my freewrite into complete sentences and thoughts. The freewrite gets polished, mapped on to the structure, and pictures from my phone are added as I go. This is where you get found out if you're using AI to write for you. I repeat: it's good editing that makes you a good writer.
6. Obviously you can ask AI to spell/grammar check as well as thesaurus.
Is that helpful? I don't think we should shy away from technology that is genuinely helpful, but there is an obvious problem if creators don't use AI to ADD VALUE to what they're doing instead of look for all kinds of shortcuts to save time.
We don't need more content in the world; we need less and better content!
Excellent write up Omar, and great as I was not able to attend! All included is insightful yet not surprising. I just feel though that we’re gonna be reading the same thing in years time. Tv measurement horses have bolted in their different directions now and not sure if incentives and legacies can ever truly align to what great effective tv measurement looks like. Device is touted but traditionals won’t let go of what is serving them in their dwindling. It’s all misaligned incentives. Thanks again (and look forward to reading the not-too-similar-I-hope next year)!
The broadcaster language is certainly changing around devices and, as Sameer politely reminded me (see below), Lantern is supposed to be an answer to this. In principle, that shouldn't be a problem. Whether you're sampling real-world audiences, or counting actual device impressions, I don't see why advertisers shouldn't expect a) to know how many people *really* saw their ad; b) what impact their adspend has on sales. Advertisers need to be more aggressive in demanding this information, not accepting what media owners and platforms say as gospel because it makes their lives easier.
Comment from Sameer on LinkedIn (complimentary of this piece but I've edited for brevity):
"The future is about fair allocation of money relative to the value created for advertisers. And on that we have nothing to fear."
"I slightly wish you’d noted that this is what Lantern is all about. We may be feeling pressure, sure, and at the same I’d probably say the response is more purposeful than panicked. Pilot is done. Beta is revving up. Transparency is absolutely going to be part of it... but it will be a different take on transparency than you need when doing counting, because incremental lift is more complicated.
"We need to skate to where the puck will be - in this instance the platforms are heading to near-term incremental outcomes, and so are we."
Full comment:
https://www.linkedin.com/feed/update/urn:li:activity:7404773198040551424?commentUrn=urn%3Ali%3Acomment%3A%28activity%3A7404773198040551424%2C7404800949526872064%29&dashCommentUrn=urn%3Ali%3Afsd_comment%3A%287404800949526872064%2Curn%3Ali%3Aactivity%3A7404773198040551424%29
In previous posts I said publishers need to be transparent about how they use AI, given that trust will become even more important in the age of AI slop and misinformation.
So, as a self-publisher, let me tell you exactly how I've used AI for this piece.
1. I used Otter.AI to record what was said at this conference. It produces a transcript which is of variable quality in an echoey event chamber, so I had to double-check the quotes I used in the article by listening back. But Otter is great at letting you follow along.
2. I pasted each transcript into ChatGPT to give me a summary of what was said. This is sort-of helpful to make sure you didn't miss anything in my notes.
**At this point, it's important to stress that during the talks, I was taking notes, but I wasn't doing the analogue strenography we're taught by the NCTJ i.e. I wrote down key points/phrases by speakers on one half of a page, but then sometimes my own reaction/thoughts/observation on the other half. So the commentary is forming in real-time, not just after the event. FWIW I use an iPad, Apple Pencil and the Notability app, which is slower than typing but I like to doodle and draw arrows.
3. The day after the event I spent the best part of an hour just freewriting what I thought about what I listened to. Freewriting is an important part of my practice and (touch wood) it has basically eliminated writer's block for me...
4. ...which is dangerous because you can produce a lot of shit writing when you get rid of your filters. A good writer is a good editor. I pasted my freeform writing into ChatGPT and asked it to suggest a structure for an essay. I didn't like the bitty format it suggested (even though it was likely more optimal for the reader who just wants to scroll and not actually read anything), but I did keep the 3 takeaways it suggested at the end. But broadly as narrative flow it didn't deviate from my freewrite much anyway.
**At this point it's important to say that the pictures are not just decorative; they are supposed to help tell the story. This is the other advantage of not being a stenographer in the room; I am able to think about taking pictures strategically. Pictures are incredibly important. I use an iPhone 17 Pro which, despite having a shitty aluminium casing which literally scratched as I unboxed it, has an INCREDIBLE camera for a smartphone (look at the panel closeups; I was 10m away in a dark room!) - tech also powered by AI.
5. Then comes the fun part: turning my freewrite into complete sentences and thoughts. The freewrite gets polished, mapped on to the structure, and pictures from my phone are added as I go. This is where you get found out if you're using AI to write for you. I repeat: it's good editing that makes you a good writer.
6. Obviously you can ask AI to spell/grammar check as well as thesaurus.
Is that helpful? I don't think we should shy away from technology that is genuinely helpful, but there is an obvious problem if creators don't use AI to ADD VALUE to what they're doing instead of look for all kinds of shortcuts to save time.
We don't need more content in the world; we need less and better content!