Jared Arnold

Media Buying, Creative Strat & DTC Growth.
Market Trader, Health Hack(er), Hopeless Surf Addict

[Jared Arnold @ads_vice]

@herrmanndigital So many Shop Plus stores had previous devs that used Script Editor specifically for checkout custom logic and had no idea that deprecated on June 30th. Big problem right now.

[Jared Arnold @ads_vice]

Been tracking HRV since 2019 with whoop. I noticed this year's ago. My brother is a functional med doc and I once bragged at how quickly my HRV and recovery would rebound after drinking. He laughed and said "holiday heart". Turns out, whoop can't tell the difference between healthy HRV and alcohol induced AFIB!

[Jared Arnold @ads_vice]

@PhilKiel Seeing the same trend! What's been interesting is taking the ads getting $0 and throwing them into a new ad set. Direct dupe, no post ID. Found some solid winners this way.

[Jared Arnold @ads_vice]

@benradack I've always been a die hard post ID fan, especially for graduating winners to a scaling camp. But in the last month or two, I've been shifting to duping. Seems to better retain the performance of the original ad I've been aiming to scale.

[Jared Arnold @ads_vice]

Anyone been hit or have a client hit with the IG account exploit? AI wild times in full effect.

From Grok... "Meta's AI account recovery assistant had a prompt injection flaw. Attackers used a VPN to match the target's region, then fed crafted prompts (sometimes with deepfake video) to trick the bot into swapping the account email to theirs and sending the reset link—bypassing 2FA. Hit high-value OG usernames hard. Meta pushed an emergency hotfix disabling the risky AI flows. No backend breach, just an overpowered LLM as confused deputy. Patched now."

[Jared Arnold @ads_vice]

@ChereneAubert Yes please!

[Jared Arnold @ads_vice]

@PhilKiel @MikeLukashewich Yeah, the angle is an FB feed (forced) + Marketplace camp. No other placements. Hyper design for marketplace. Feed will do ok, but MP will rip. Pro tip... Add a red or other color stroke border on the static. Minimal to no overlays or text.

[Jared Arnold @ads_vice]

@HireFireTeam Over what time frame does this data cover? I ask because FB Reels has been pushing into feeds more and more over time as older audiences tend to not click into the reels environment itself.

[Jared Arnold @ads_vice]

@binghott I didn't. Honestly, I was happy to get out of there. I run a lean stack of clients and love working with all of them. This one was a pain across the board.

[Jared Arnold @ads_vice]

@binghott No, the same DNA that had them thinking that was ok trickled down into many other business decisions. I ended the engagement swiftly once we closed out the year.

[Jared Arnold @ads_vice]

@binghott Man, it was the most astounding thing I'd ever encountered. Short lived engagement you can imagine.

[Jared Arnold @ads_vice]

I now exclude this as a rule with few exceptions across all clients. The value that placement provides was just not worth the risk of overspend on them on a random basis. Tough to say to what degree it contributes to performance overall but it definitely didn't hurt (which says a lot). Couple that with the "sleep at night" factor of not having it eligible and it's a win/win from my perspective.

[Jared Arnold @ads_vice]

@PhilKiel Underrated strategy

[Jared Arnold @ads_vice]

@dtcprophet Even if it's wrong, it's awesome!

[Jared Arnold @ads_vice]

Yeah exactly, only in platform with no 3rd party incrementality measurement. And I totally remember that post of yours! It was actually on my mind as we ran some of these A/Bs. I don't like to over-conclude anything from in platform data and these tests. Just found it funny that targeting the objective didn't always result in achieving that objective. We don't see that in the funnel events the same way. If I target ATCs as an obj., I'll get them efficiently but purchases will suffer. If I target purchases I might have higher cost / ATC but I don't care. That's how it should work. Another confounder is the scale of the client. Smaller, early stage brands have an incrementality advantage because they're just not out there and nobody is aware of them. Almost all of your paid efforts are by default, incremental.

[Jared Arnold @ads_vice]

@binghott @IstvanicMarin Meaning, sometimes, we'll see higher IA ROAS, lower IA CPA, etc. Doesn't speak to whether or not in-platform IA is accurate in general. Just that you would expect an IA objective would result in better IA metrics holding other variables constant (like audience, creative, etc.).

[Jared Arnold @ads_vice]

It's a great app experience, seamless local blood draw scheduling, and a very solid panel of biomarkers a GP and insurance would never order. Try to a la carte that biomarker list with ulta or quest and it'll be 2 to 3x. And that doesn't even account for a partial 6 month retest that's all part of your fee. Personally, depending on any specific risk factors you know of or issues you need a deep dive into (autoimmune, CVD risk, etc) plan on a spending a little extra on some of the add ons beyond the baseline panel. The deeper lipid panel is really useful and worth getting as an example. And I think they've added a lot more options since I did it last.

[Jared Arnold @ads_vice]

@binghott Investing and media buying are the only two disciplines I know where doing more does not equal better performance. And both have clients that don't know that. Hah!