top of page

A/B Testing with AI Tools to find Message Market Fit

Updated: Jul 20



baby Fist Pumping with Email Stats


I have been buried for the last week doing nothing but A/B testing with Clay. It has taken a week of constant testing with small batches of very targeted emails to hit my metrics.


I ran a slew of enrichments and filtered on multiple criteria while building these lists, then I split the list into 3 batches for testing using a formula in Clay.


Linkedin and Clay Search Filters
Filters for LinkedIn and Clay

Email enrichments with Clay
Enrich with emails

Now I know whether the company is the right size, whether they are looking for more clients, whether they have help for outbound sales activity and other criteria that would determine if they are better fit for this offer.


LinkedIn lead filters
LinkedIn Lead Fliters

Clay AB Test formula
Using a formula in Clay to split tables for AB Testing

By being laser focused on my ICP I can quickly adjust my outgoing messaging to hone my metrics.

You can use AI to help craft your initial messages and to even summarize your landing page offers to ensure the messaging remains consistent throughout your campaigns.


Clay Agent Summary
Clay Agent Summary of Offer Page

I know that my current sample size is small, but if it follows the patterns of my previous tests, I am confident that I am getting closer to my message market fit.


Twain AI Email
Twain AI

If you are adopting AI tools to help you scale and increase your engagements to your ICP’s, please do not just grab a cookie cutter template that you see on LinkedIn and run with it.

Nothing convinces a potential client that this message is spam more than receiving 10 other emails with similar messaging.


This is one of the problems with all the AI tools I have been testing for outreach.

They all seem to follow that same formula and offer no insights into how you can refine that offer to better target your ICP’s.


The only way to really refine that offer and understand what your targets are looking for is by testing, testing and more testing.


AB Tests
AB Test Results

This is a lot of work but once you have done the initial testing and have a response rate you are satisfied with.

That is when you can start to scale that offer out to multiple channels.


Once I know people are intrigued by the Subject and that content resonates enough. I can start to work on the offer and increasing my replies.


The last 2 pieces would be my offer refinement and followup sequences.


If you wanted to learn more or get a free sample table, head over to our ICP Profiler page and signup for it there.


Comments


bottom of page