- Chew On This
- Posts
- 🥗 Irresistible Offers + AI = The Perfect Cocktail
🥗 Irresistible Offers + AI = The Perfect Cocktail

Hey there
Welcome back for another bite to chew on.
Hope you’re having a great weekend with your loved ones and are ready to crush the upcoming week
In this weeks newsletter, we’re talking about two things;
A) How to craft the perfect offer for your brand
B) How to use AI to supercharge your marketing & creative department
This will be one of those jam-packed newsletters, filled with tactical gold that you can literally implement in your business, today.
But before we dive into it - we have two things we want to share with you…
1). Hundreds of people love our “Daily Chews” SMS service where we’re texting you daily nuggets of valuable information.
We’re sure you’ll love it as well. If you want to subscribe - simply go here!
2). On Thursday - we’re releasing a brand new podcast series called “Chew On CRO” where we’ll teach you everything you need to know about CRO. We can’t wait to show you what we’ve been cooking up the last few months!
Now - let’s get into today’s newsletter
Understanding your numbers to craft the perfect offer
First things first - let’s talk numbers.
Without understanding your numbers, you can’t do anything.
You can’t make irresistible offers, and you certainly can’t run ads to that offer (profitably)
So - let’ 's start by understanding the most foundational piece in offer creation.
Numbers

The two “north star” metrics you need to know by heart are:
Contribution Margin
&
Breakeven ROAS
If I wake you up at 3am in the morning and ask what these two metrics are for any given offer of yours, you should be able to answer right away.
The calculations are pretty simple:
MSRP (Manufacturer’s Suggested Retail Price) - COGS - Discounts - Shipping - Processing Fees - CAC = Contribution margin
MSRP (Manufacturer’s Suggested Retail Price) / Net Profit Before CAC = Breakeven ROAS
Now, look;
The reason why we say you need to know these two metrics by heart, is because we’ve made the mistake of NOT knowing these numbers by heart - and we don’t want you to make the same mistake.
For a long time - we were crafting offers with free this and discounted that, but what we didn’t think about was it increased the break-even ROAS to a point where that offer became almost impossible to scale.
So;
Learn from our mistakes. Understand your numbers.
Next up -
Let’s take a look at what a BAD offer looks like, and what a GOOD offer looks like (real example from Obvi)

Good offer vs. Bad Offer
First of all - the goal of this offer was to increase our AOV and incentivize buying a bundle of 3 to 6 months of supply for our collagenic burn
Now - let’s break down what the right and wrong approach was to achieve this goal
Bad Offer:
So, the way we increased our AOV (in the bad offer) was bundling & offering free gifts and additional discounts with each bundle increase.
But…
The issue with this was
We knew that we were increasing our COGS.
But at the same time, we were also thinking that this offer would improve our Conversion Rate enough to cover the additional COGS.
Buuuuuut…
It didn’t
And tbh - we didn’t really monitor it until waaaay too late.
I.e. - the lesson learned is; Do the opposite of what we did. And always do monitor what you’re doing.
Besides that - the “bad offer” also needed a 2x ROAS in order to breakeven - which in and of itself just makes it too hard to scale.
Our Solution:
After taking a hard look at these numbers - we essentially re-crafted our offer - removed all the free gifts and increased the prices where we could.
As you can see, our “Good Offer” only consisted of additional discounts with each bundle increase and eliminated all freebies.
This brought down our breakeven ROAS to 1.44x.
To determine a winner, we A/B tested this offer and surprisingly our conversion rate didn’t drop enough to completely ignore this offer.
Let us explain…
As you can see there was a difference of around 28% between the two breakeven ROAS.
So, our conversion rate did drop (obviously) but the drop was only up to a 28%, making the second offer much more appealing (from all perspectives).
Long story short - make sure to calculate your breakeven ROAS when crafting your offers because as low as you can get it the easier it will be to scale that offer.
4 Offers you should be testing
Whether you’re a newly founded brand, or a mature brand with multiple millions behind you - you should ALWAYS be testing new types of offers as well as new combinations of offers
Here are some of the offers that have worked well for us at Obvi , which we recommend you giving a shot
1) Bundling
I.e. Offering more of the same product at a discounted price. (the one we showed above).
2) BOGO Offers
This is basically also a bundle offer, but just framed differently
3) Cash Back offers
Here’s an example from a brand that’s also crushing it online

4). Gifts with purchase
“Buy for X amount” and get a free mystery gift. This works great as well - and is adding another level of excitement to the offer.

Pro Tip: Automatically apply discounts instead of using discount codes or the like. They just add an extra step of hassle and ultimately hurt your conversion rate
How to use AI to boost your creative process
We see a lot of tweets about using ChatGPT for copywriting but most of them lack a ton of structure as to HOW you should effectively prompt the machine.
You see, for any AI algorithm to spit out good outputs - you need to give it good inputs.
… and to consistently give it good inputs - you need a solid structure to follow
Here are the 4 primary things we use AI for, in relation to creative development at Obvi:
1). Review mining and understanding consumer behaviour.
2). Writing ad copy from the insights generated above
3). Writing landing page and advertorial copy
4). Writing UGC briefs
Now - let’s break down each of these
1). Save hours on doing customer research:
At Obvi we use ChatGPT to mine all of our Amazon and website reviews to get deeper insights into our customers so we can speak directly to them.
Here’s how we do it:
a). Use an Amazon scraper/ extractor to give you a CSV of ALL your Amazon reviews + 1-2 of your competitor’s reviews.
b). Take the CSV and put it into a Google sheet, clean it up and make a PDF you can upload to ChatGPT using the AskYourPDF plugin.
c). Get AI to read through the entire PDF.
d). Now you’re set for prompting and asking it questions. Make them specific to your customer like understanding their pain points, their desired dream outcomes, their objections etc
Here’s a specific checklist we follow:

2). Writing Advertorials you can run ads to
Advertorials and listicles are a MUST HAVE in any DTC marketers arsenal.
But…
The problem most people have when it comes to advertorials and listicles, is the amount of time it takes to write them.
Luckily, there’s a way to get ChatGPT to help you get started with the process
Here’s how we do it:
a). Tell the AI who it should be. For ex, a world-class copywriter.
b). Tell what it’s supposed to do. For example: Write an advertorial that will be featured as an advertisement.
c). Tell the tone it should write. Ex: be persuasive, clear and concise.
d). Tell where it should get the info from. Ex: The reviews + the insights from use case 1 AND tell it how to structure the article (the framework to be followed).
Pro tip: Use prompt-genie.com to compile these super prompts for you ($3 per month)
Important note: ChatGPT can not (and will probably never) be able to replace an excellent copywriter, who has a deep understanding of direct response marketing. But - it can definitely help get the creative juices flowing, create first drafts, and ultimately: Save a ton of time
3). Writing Ad Copy
Now, this is exactly the same as writing an advertorial which we covered in use case 2.
The ONLY thing that changes is step #2 where we ask the AI to write ad copy for a Meta Ad in XYZ format {Headline, Primary text, description}
4). Writing UGC briefs
Now, coming up with UGC briefs is a little hard because we have to make it sound authentic and not too robotic.
Here’s the skeleton we follow:
a). Use the reviews to come up with a list of the following: pain points, desired outcomes, purchase triggers, unique value proposition, uncertainty and perceived risk, and objections - based on the LIST of reviews.
b). Follow a “performance creative” framework, similar to the ones listed below. (There are many more of these out there - but these are just some of the ones we’ve used successfully in the past)

c). Iterate on each “module” of the creative - by adapting the following “parameters”:
Verbiage
Readability
Price Wording/Framing
Tonality and Avatar.
d). Compile the script by putting together each seperate “module”
Tool of the week
Paying for contacts who you’re not contacting sucks…
Are you tired of spending money just to host your email lists - even though you don’t send them your emails?
So were we…
That’s ONE of the main reasons for why we switched to Sendlane.
Secondly, their support is second to none. They have a 24/7 Slack channel where you can bring up any issues (if they arise) and get them resolved ASAP.
Thirdly, their CEO Jimmy Kim is a great guy and you can directly tweet at him to request new features that should be added to the platform
… and the best part is: They even get it done pretty damn fast.
If you also think paying for contacts absolutely sucks…
And…
You feel like talking to Klaviyo Support is like talking to a wall…
Then we definitely think you should consider switching to Sendlane
#ProudPartner
Thanks for reading along
That’s all we had for this week’s newsletter
As always, we appreciate you taking the time to read this. We hope we were able to provide some value to you and look forward to serving you again on Wednesday.
All the best,
Ron and Ash