AI has been hyped up for a while now, but it’s starting to get real. Especially in GTM.

Unfortunately, most organizations have only scratched the surface. They’ve given their team licenses to ChatGPT or other platforms and very minimal training, and/or tried to implement an off-the-shelf solution with little to no impact.

There are few stories of GTM teams implementing AI and seeing real impact.

Those stories are starting to come out. Jason Lemkin at SaaStr recently shared how they successfully built an AI SDR that outperforms their best human SDR. Part of this is the tech getting better. Most of it is the GTM team getting better at leveraging that tech.

In this newsletter, we’re going to share some of the best use cases for AI in GTM and how teams can actually deploy AI to move the needle on revenue production. It’s still early days, so we’re going to use first principles to look at what works and doesn’t work with humans to think through how to deploy AI successfully, just as SaaStr did.

The Dream: AI Marketers, SDRs, AEs, AMs and CSMs

Let’s start with the fantasy.

In a perfect world, AI runs our entire GTM motion from top to bottom:

• AI Marketers and SDRs fill the pipeline with high-quality leads
• AI AEs close deals faster than our top producing human reps
• AI AMs and CSMs retain and expand our customers for us

Today, that fantasy feels about as far-fetched as simply asking ChatGPT to “make me a billionaire”, but there are components of each that make sense if we break them down.

Last year, a lot of companies tested out fully autonomous AI SDRs—and we all received a lot of terrible sales emails as a result. SaaStr recently made an AI SDR work by NOT making it autonomous and working alongside it.

Let’s dive into each of these areas and break it down.

What Actually Works (and What Failed)

If you read Jason’s article, he makes a big deal about the foundation they needed to build to make their AI SDR work.

SaaStr had:

• A world-class brand
• A warm list of opted-in contacts
• Clearly defined ICPs and Personas
• Refined messaging and proven outbound playbooks
• Humans reviewing and coaching on every AI message

In other words, they built a repeatable outbound motion first, and then used AI to scale it.

Meanwhile, most of the AI SDRs that failed did so for the same reasons human SDR teams fail:

• Vague ICP
• Poor targeting
• Generic messaging
• No process, no coaching, no accountability

This produced a lot of the same spray-and-pray messaging that many human SDR teams were failing with. They used AI to insert company names, titles, etc without making the messaging any more relevant to the buyer. We all received those emails and ignored them.

Foundational Elements

Before we can ask AI to help us execute, we need to build the foundation. Here are some of the basic elements of that foundation and how AI can help us.

Clear ICPs & Personas

Like humans, we have to point AI at the right buyers. This literally defines what buyers are easiest to sell so, by definition, if we don’t define this carefully we will be pursuing prospects that are more difficult and convert less.

AI can help us clearly define our ICP and Buyer Personas, if we arm it with the right data. We can aggregate our company data from our CRM, emails, call recordings etc. and even augment that data with other tools to feed the AI deep intelligence on our best and worst customers and sales opportunities.

Target Account Selection

Once we’ve aggregated all our data and defined the ICP, we can then use AI to help identify the best accounts to sell to, both prospects and existing customers. Assuming we have the right data on both, from third-party data to product usage, we can have AI help score accounts to identify the best.

We talked a lot about the value in doing this in our Outbound Framework and Allbound Framework.

Lead Scoring

Overlapping with target account selection, we can use this same data to help score leads and/or accounts that came in from, or were influenced by, our marketing engine. AI should be able to help us compare the deals we’ve won and lost and our best and worst customers to the new leads we have coming in.

GTM Process Mapping

Those gold leads and accounts above aren’t worth much if we don’t follow the right steps to convert them. If we have a lot of data from call recordings, emails, and CRM data, AI can use it to help us identify the critical steps in our process as well as to monitor the execution of that process in future emails, call recordings, and CRM updates.

These four sections apply to everything from new business sales to marketing to customer success. By better defining our buyers, identifying the right accounts and leads, and mapping the right process, we’ve set the foundation for GTM.

More on this in our GTM Efficiency Pyramid Framework.

Sub-Foundational Elements

Okay, so let me warn you this section is going to be boring. It should come before the above but I didn’t want to lose you.

We can’t do much of the above without even more fundamental elements.

Segmented, Accurate Data

Most companies have bad data, and feeding this into the AI is going to produce poor results. SaaStr had to do a big data clean up exercise to make their AI SDR work. They had deals missing from their CRM and many other gaps in their data that would have led the AI to form the wrong conclusions.

The CRM and Marketing Automation tools need to be enriched, cleaned, and segmented before AI can generate meaningful output. This takes a lot of work. Unfortunately, there’s no way around it. (Though our team can help here.)

Human in the Loop

AI can add a lot of value to every step in GTM but there needs to be a human in the loop. We need to feed it instructions, review the response, provide additional feedback, and do revisions to get it right. None of this is a set it and forget it easy button.

Driving Revenue

Okay, with that foundation in place, let’s get into the good stuff, where AI can start to move the needle.

Pipeline Generation (AI Marketer and SDR)

• Help draft content
• Research accounts and contacts
• Write highly personalized outbound emails
• Score and prioritize leads based on intent + fit
• Aggregate data to identify whole accounts to prioritize

*But again, it’s important to set the foundation and have a human in the loop.

Pipeline Management & Closing (AI AE)

• Summarize call transcripts and flag risks
• Suggest next steps based on historic patterns
• Auto-update certain opportunity fields in the CRM
• Draft follow-ups, recap emails, help prep for sales calls

*I don’t believe AI can run sales calls today, or even move deals in the pipeline from stage to stage, but with a human in the loop, it can help a lot with research, call prep, updating the CRM, and follow up.

Retention & Expansion (AI CS)

• Analyze product usage
• Review customer interactions
• Identify healthy and at-risk accounts
• Draft proactive CS outreach messages
• Help prepare QBRs, Account reviews, etc
• Surface expansion opportunities to CSMs/AMs

*There’s always a lot to unpack in CS but customers offer a lot more data and patterns for the AI to recognize that prospects, making the use cases really compelling, if the right humans are in the loop and the right foundation is in place.

Common Traps to Avoid

Even with the right tools, many teams sabotage their AI initiatives by:

Skipping the foundation: No ICP, no defined process, no coaching = AI accelerates chaos.

Trusting AI outputs blindly: Always review and validate. Use AI as a co-pilot, not a replacement.

Failing to segment: You can’t use the same messaging for an enterprise buyer in finance and a mid-market ops leader in SaaS. Train the AI accordingly.

Over-automating outreach: If you wouldn’t send the message yourself, don’t let AI send it.

Final Thoughts: AI Is an Amplifier, Not a Savior

If your GTM engine is broken, AI won’t fix it.
If your GTM engine is solid, AI can help you 10x it.

The best use cases for AI aren’t about removing humans.

They’re about removing grunt work—so your humans can do their best work.

• Better prep
• Better results
• Better coaching
• Better personalization

Just don’t skip the hard stuff.

Want to Make This Real?

We’re helping B2B SaaS companies build the foundation they need to turn AI into a force multiplier across the GTM funnel.

• Need help mapping your process?
• Want to get your ICP and personas cleaned up?
• Curious where AI can (eventually) 10x your output?

Let’s talk.