This is my favorite way to test offers quickly, so you have the most rapid feedback and can figure out what to scale.
We use it for cold email but it can work on any channel as a framework. https://x.com/i/article/2115462078177611778
The Gauntlet: How We Test Cold Email Offers Without the Fluff
We took over a campaign where the company's internal team had generated 3 leads in 2 months.
We generated 85 leads and 4 closed deals in 2 weeks.
The messaging framework that got us there came out of Growth Engine X, the agency I built, which generates 300-400 positive responses a day across its customers.
With it we learned things like,
1. Offering 2 free meetings on someone’s calendar works better than launching a free campaign for them or even sending them every contact in their full market with valid emails.
1. No one wants to talk about bookkeeping but they would have an accountant review their current tax plan to find things their current accountant missed.
1. Restaurants care about increasing their events revenue but don’t care about their reservation cancellations as much.
Internally, we call it "The Gauntlet."
The one-line email test
I wanted a way to test offers quickly and know exactly what was working and what wasn't, without the fluff.
So we started testing one-line emails. Literally something as simple as:
"If we could help you {{offer}} without {{risk}}, would that be useful?"
or
"We just helped {{case study}} {{accomplishment}}. Could I show you how we did it?"
What I love about the one-line email is that if they don't respond, nothing else could have gone wrong. We didn't mess up the first line. We didn't screw up the social proof.
For this framework, assume the list is correct. If the list is good and nobody responds, you get clear feedback: the offer and the next step didn't work.
How we run it
We test two things at once, at two different levels:
1. The angle changes dramatically from bucket to bucket.
1. The offer changes within each bucket.
We write 3-5 angles per bucket, across 3 buckets of friction:
- Bucket 1: what we want them to say yes to
- Bucket 2: what we want them to say yes to, but it takes more work on our side
- Bucket 3: what we really don't want them to say yes to, but it shows how far our offer has to go to get a yes
How we judge it. Positive replies are the win for each bucket. They're the leading indicator for meetings booked, and meetings booked are what matter most. In the end, the real winner is whichever campaign brings in the most revenue.
How long we run it. We send each sequence to 1,000 people before we call it quits.
Pro tip: people don't remember the emails you send them. If you want to test different buckets on the same contacts, wait 7 days between them and change the subject line. You can test every bucket against the same pool of people.
Bucket 1: drop them straight into your sales process
Say we're selling lead gen services like Growth Engine X. Ideally, we'd email someone "Hey, do you need more leads from channels like cold outbound?" and they'd say yes.
That would be a dream. Of course, it doesn't work for us.
That doesn't mean it won't work for others. When we worked with Numeral, we literally asked people "Do you need help automating your sales tax compliance?" and got a ton of great responses.
Bucket 1 tests messaging that drops people right into your sales process.
Bucket 2: more work for you, still in your sales process
This bucket adds work on our team's side but still gets people into our sales process.
At Growth Engine X, we offered (and still offer) cold prospects 2 free leads booked directly on their sales calendar. It's a lot of work on our side, yes, but we got far more positive responses with this offer.
If you do something like this, put a qualifier somewhere else in the message so you don't attract the completely wrong leads. We used the P.S.:
"P.S. Of course these leads are free, to show you we can get meetings. If we're successful, we'll talk about putting you on a flat monthly retainer like the rest of our customers."
That tells anyone who thought we'd always work on performance, up front, that we don't. For some reason those people still respond and you have to weed them out in your sales process, but alas, we keep moving.
The P.S. is part of the test, too. It changes who replies, so treat it as a variable, not decoration.
Bucket 2 tests offers that get people into your sales process but take some free work or product to grease the wheels.
Pro tip: this is where we test lead magnets. DO NOT build a full AI agent to fulfill your lead magnet, or a process on your team, until you know people want it.
You can always turn the campaign off and take 24-48 hours to build the AI, system, or team process to fulfill it.
I warn you because I've seen people spend engineering resources on full-blown lead magnets nobody wants. Testing first and building after costs almost nothing.
Bucket 3: the offer you hope you don't need
This is where we test the things we really hope aren't true. We have to know what responds, and who, when we offer things we don't actually want to do.
This bucket cuts through the noise when you think you have a cold traffic offer but really don't.
It's where we try paying people to take meetings, doing a month of free work, or giving away something very valuable we don't want to do over and over.
If you're afraid to test this, test buckets 1 and 2 first. If you get nothing, take something from bucket 1 and offer people $50-$200 to take a meeting. If you can't get them to take a meeting after offering them cash, you have undeniable proof you don't have an offer that converts cold traffic.
If bucket 3 wins, you're not done. First, check whether it works with your cost to acquire a customer. Then go back and rethink what you can offer in buckets 1 and 2. Use what you learned to figure out how far you really need to go.
What this looks like in practice
Back to that commercial waste company from the top.
First, we outlined what buyers of their service actually find valuable. It came down to four things:
1. Winning on price
1. The driver is always the same
1. The actual invoice matches the original quote
1. Compliance with state regulations, backed by an insurance guarantee other companies don't offer
Those became the angles. We tested three of them inside Bucket 1, all ending with the same ask: "Could we talk about how our service could fit your needs better?"
- Same driver, same day: "A lot of offices we talk to are walking a new driver through every few weeks." 6,111 people contacted, 5.38% reply rate, 2.08% interested.
- Invoice matches the quote: "We hear a lot about a box rate that turns into a much bigger bill once fuel and fees get added." 10,450 contacted, 2.10% reply rate, 1.23% interested.
- Compliance walkthrough: a free on-site review of where state rules actually apply. 3,985 contacted, 3.24% reply rate, 1.12% interested.
The same-driver angle shockingly got more interest compared to invoice-vs-quote. People don't hate their bill as much as they hate retraining a stranger every few weeks.
Then the offer changes across the buckets:
- Bucket 1: ask if they'll take a meeting to review what working with us could look like.
- Bucket 2: offer to pay for the next 3 months of their contract with their current vendor.
- Bucket 3: buy out the remaining term of their current contract, up to a full year. (This industry normally runs 3-year terms, so that's a real number.)
In this industry, the biggest reason not to switch isn't the pitch. It's being locked into a contract. Buckets 2 and 3 tell you exactly how much of that lock-in you have to remove before people say yes, and whether that still works with your cost to acquire a customer.
(If you want to read another article about converting cold traffic, I'll post it soon.)
You’ll actually notice that we don’t have metrics for Buckets 2 and 3. In this case, we got enough leads from just asking for a meeting that we didn’t need to add the extra cost and friction to the sales team in order to get positive responses.
But if Bucket 1 didn’t work, we could of course make the offer better and better.
If you’ve read this far, give this article to your AI agent to help think through your offer and how you can test multiple variations of your offer and what it takes to get a positive response.