Peter
@Peter_Quadrel
Followers
1K
Following
4K
Statuses
5K
Incremental growth for premium & luxury brands. At the intersection of AI advertising & finance.
Spreadsheets
Joined September 2022
There's only two types of D2C brands: 1.) Aesthetic/Pleasure Based - Fashion - Home goods - Cosmetics - Some Food/Drink 2.) Solution Based - Skincare - Electronics - Automotive - Supplements If your brand falls under 1 you're a "I see it, I like it" brand. Your job is to test as many aesthetic designs, varieties, flavors etc. as possible, NOT to test an extreme amount of angles or formats. There may be some angle testing, but this is secondary to sku variety and creative testing itself. If your brand falls under 2 chances are you're a "Problem solver" brand. Your job is to test as many problem, solution, benefit, proof angles as possible. Aesthetics and variety of sku are secondary here, focus all on the concept. Some straddle this line. Even then audience segments always lean one way or the other. Develop assets accordingly.
19
18
337
@andrewjfaris @binghott Do a correl study on your data set, varies by brand but def worth the time it’s the only set of video soft metrics I’ve seen with this relationship
1
0
1
@andrewjfaris @binghott Watch time/thruplay is pretty correlated to spend and conversions as well
1
0
1
How to actually vet an ad agency: 1. Ignore case studies probably old and inflated 2. Contact clients past and present 3. Check Glassdoor unhappy employees expose the reality 4. Get an audit and focus on their thought processes and system 5. Check legal database for past lawsuits with clients or employees 6. Hire in house
0
0
3
@thesamparr GX or NX will last the longest, sporty enough for dad enough space for mom and easy to drive for gma
0
0
0
Lack of CLARITY is why most ads fail. People literally have no idea what you sell, how it works, or why it matters. Show the thing and/or the results so blatantly that your grandma can SEE IT and UNDERSTAND IT. A bold headline with product on white beats an army of designers trying too hard most days.
0
0
2
@TaylorHoliday @MikeLukashewich @codyplof @londonspilker Would at least need stat sig reach volume relative to Meta TAM. But doesn't really work on bayesian
1
0
0