Tweet épinglé
Shiv
9.1K posts
Shiv
@ShivEcom
brand owner & ecom operator. i scale my own brands and Emails + CRO for You
DM Inscrit le Ağustos 2016
338 Abonnements804 Abonnés
Shiv retweeté
most ai ad creatives don’t fail because the product is bad
they fail because the prompt is garbage
brands ask claude for “facebook ad copy” and get the same generic slop everyone else gets
so i built a system for writing ad creatives that actually sound like a real buyer wrote them
it breaks down:
- the product
- the customer
- the pain
- the desire
- the objection
- the angle
- the ad structure
- the testing variation
then claude turns that into usable ad copy instead of robotic filler
the goal isn’t more ai content
the goal is better ads, faster testing, and copy that feels specific to the customer
for the next 48 hours i’m giving it away:
- the claude ad creative workflow
- the prompt structure
- the copy framework
- the worked example
- how to adapt it to your own product
rt + comment "copy" and i’ll send you the claude ad creative system
(follow for dm)
English
most brands don’t need more products
they need more ways to sell the same product
this is the workflow:
claude takes one product and turns it into multiple ad angles
gethookd shows what winning ads already look like inside meta
then you stop guessing and start building from proven patterns
same product
different market
different desire
different hook
different visual frame
pet brush
sparkling drink
mini vacuum
3 completely different ads built from the same system
that’s how you increase creative output without turning every ad into random ai slop
rt + comment "ideas" and i’ll send you the claude + gethookd ad workflow
(follow for dm)
English
If you haven’t added Google ads to the mix, you’re missing out
English
THE GAME OF GOOGLE ADS
LEVEL 1
> Set up conversion tracking
> One Pmax campaign
> A few conversions
> Feed directly imported from Shopify
> Non-brand search with a few generic keywords
LEVEL 2
> Brand separated from non-brand
> Goes through the searches with impressions and realizes it's a lot of trash
> Built up around 50+ conversions
> Hits a decent ROAS
> Better feed titles added via a supplemental source
> Can't scale past $10K per month
LEVEL 3
> Able to start splitting products up into more than 1 campaign
> Works with custom labels to split products
> Feed optimized deeply via Channable, Datafeedwatch, or similar
> Non-brand search built out into multiple angled campaigns
LEVEL 4
> Never limited by budget
> Has YouTube as the top spending channel
> Fully dynamic label shopping setup
> No Pmax
> Search setup targeting different awareness stages and using contextualised landing pages for the layer
What level are you playing at right now?
English
Raw image vs TryFrame
No more guessing - you can now try on your glasses with AI.
The Best Designer@Chinwe_Owanta
TryFrame is an app that allows you to virtually try on any pair of glasses - in order to see how they fit before you make a purchase. Try it here: app.flora.ai/techniques/try…
English
TryFrame is an app that allows you to virtually try on any pair of glasses - in order to see how they fit before you make a purchase.
Try it here: app.flora.ai/techniques/try…
The Best Designer@Chinwe_Owanta
My pair of glasses snapped on camp - so when I got back to Lagos, I went to pick up new ones. This time, I got two new frames - each complete with their lenses. On getting home, I realized they weren’t really my style - yet they were overpriced. So I built a solution.
English
Here's how I validate the product B4 branding it.
Only 5 things needed:
-Meta acc
-3X budget of the product
-1$ shopify
-30 to 40 ads minimum
-Ripped decent branding
-Metrics in acc
You build 2 funnels:
1 mid problem to solution aware:
You relate them on what they are pain points, experiencing, dealing, going thru then you take them to your solution. Give them a good offer, done.
1 solution to product aware:
You look at the current states of products in the market, do claude if you can make the winner product better or nah (this is really important) then you start running comparison ads of why your product is better. (has less carbs, has more protein or is easy to use etc)
now you go ahead and make the ads (min 30)
You put all of them in ABO, let them run for 3 to 4 days and Don't look at it (search more products at the time)
after 3 days you look at the metrics
CPM: If it's high, that means market has less people who would buy it. if It's low GGs now look at CTR.
CTR: If CTR is low it has low demand thus I usually stop unless I wanna run it up so I do Unaware less that lead to problem and solution. if CTR is high good.
CVR: if in these 3 days lets say you spend $150 and you break even: zero thoughts run that product
If you made half of it back: I'd make few more ads and push CBO with best metric ads.
if there's zero sale, FUCK IT WE BALL change the product.
that's my method.
English
Hired this new intern to design with me but he keeps sleeping mid work 🤦
English