How Jordan Crawford’s outbound sales method is helping GTM teams

In the era of AI slop and personalized yet impersonal mass emails, outbound sales can hit many hurdles attempting to break through the noise. Jordan Crawford—a fractional GTM engineer who founded consultancy firm Blueprint GTM and writes the Substack On the Edge—has created his own B2B sales method to address these barriers by combining AI and public data to create more targeted and personalized outreach.

Revenue Brew sat down with Crawford to discuss his B2B sales method and how it can be applied to various industries.

This interview has been condensed and edited for clarity.

How did you come up with this method?

The core premise of my argument is the way most people are thinking about go-to-market is they first define their ideal customer profile—and that usually looks like it’s B2B SaaS businesses between 50 and 100 employees located in North America—and then they go and define their personas. “I care about the VP of marketing and the VP of sales.” Then they hand a list over to a BDR. Basically they get a message that is every shit message that you’ve ever received: “Hi Jordan, I see you’re in B2B tech sales, too. You wear a hoodie. I wear a hoodie too. We’re the same person. Can I just get 15 minutes of your time to talk about how my AI B2B SaaS solution is the best ever?”

It’s not really a money problem. It’s that AI has commoditized messages that weren’t really valuable in the first place.

How does AI fit in with your sales method?

People are trying to use AI and deploy a bunch of tokens at a company and find out everything about this company, and know everything about me, and then write a really great message. There’s no theory in that and AI doesn’t have any guiding principle. My process is a little bit different. It generally works really well for vertical SaaS businesses. Horizontal SaaS businesses can do this, but they have to pick a vertical to focus on.

In my methodology, the very first place to begin is to create these things called account dossiers, and it’s everything in the CRM, everything [customers have] ever said, and everything they’ve ever done inside of your app. You have a full timeline of every single customer and then you ask Claude: “Go look at my closed lost deals, my closed won deals, and then the customers that are just doing bangers.”

Then go rewind time to when they came to us. How could we have determined they were in that situation ahead of time? It turns out that there’s 100 times more public data and Claude is amazingly great at reasoning backward…It can do a holdout test. Here’s a way to identify a 10x customer just using public data, and once you do that, you can target by situation or by pain.

What are some examples of using public data?

You can deploy one of two messages. The first message I call a PQS, a pain qualified segment. All you’re doing is mirroring that prospect’s situation back to them based on the situation your last customer’s in. [For example,] a pool cleaner company, once you add your first pool tech, my guess is that you’re having trouble managing your scheduling. I read some Google reviews that seem to [say] it’s hard to manage scheduling now that you’ve added your first pool tech. That would be an example of a pain qualified segment. But there is a much better message, which is called a PVP, a permissionless value prop, and that is a message that is independently valuable to the person receiving it.

Does this method work best with clients in certain industries?

My methodology works really well for vertical SaaS companies ideally in heavily regulated industries where there’s really great public data. I’ve worked with companies that sell into the junk hauling space, that sell to fire departments. I’m working with a bunch of companies that sell to dentists right now. It turns out there’s a lot of really great public data to identify. You can look at Google reviews. You can look at licenses per state…This works best where there’s a combination of clear pains that you solve, and where you understand your customers very well. There’s way more public data than you think when you let Claude go down a rabbit hole.

What are some best practices you advise for striking the right balance between AI and the human seller?

Empathy is not dead and this process works best by having greater empathy and understanding of the customer, and using AI to amplify that empathy, not replace it. AI’s job is to remove more of the mundane things that sellers do, so they can get back on the phone, so they can get back in person, so they can talk to customers.

If you’re buying a CRM to run your pool cleaning business or dental software or dental education, those are considered purchases. You want to know the person who you’re buying those things from. You’re running your business on this thing and that’s where you want to have a conversation with a person. AI’s job is more to figure out what accounts in your [total addressable market] you should never talk to…If you can score and sort your market and say this is the top 10% of my market, the question is: What can you do to create an exceptionally memorable experience?