How B2B marketers (actually) use AI

What’s Your AI Tool Stack? Real Marketers Share What’s Working Right Now.

Editor’s Note: Hey. Dave again. This question keeps coming up: what’s everyone’s AI tool stack? And I get it. There’s this fear of missing out. Like everyone else is using these amazing tools and you’re stuck just using basic ChatGPT. So we asked marketing leaders to share what’s working for them right now. And it turns out most people are still using the basics. ChatGPT. Claude. Maybe one or two other tools for specific problems. Nobody’s building some massive AI operation. They’re just solving annoying problems in their workflow.

The question isn’t what tools you should be using.

It’s what problems are you trying to solve.

Sales reps wasting time hunting down answers? There’s a tool for that. Landing pages that feel generic? Someone’s figured it out. CRM data that’s always incomplete? Fixed.

This isn’t about having the perfect stack. It’s about finding one or two things that remove friction from your day.

Here’s what’s working for real marketing leaders right now.

💬 The Custom Sales Chatbot That Saves Hours Every Week

One marketer built a custom sales enablement chatbot with their fractional rev ops team.

Their product sits in an ecosystem with other dependencies. Sales reps constantly field questions about which vendors to use, which integrations work best, who to recommend.

The old way? Ping someone in Slack. Wait. Hope they’re available.

The new way? Sales reps ask the chatbot during live customer calls and get answers immediately.

They trained it on their internal directory, resources, and vendor information. The chatbot sits in the background. When a customer asks a tricky question, the rep types it in and gets the answer without breaking stride.

The result? Sales is getting 80% of what they need in real time on calls.

Why not just use a custom GPT? Dataset limits. Custom GPTs cap how much information you can feed them. Building their own removed those restrictions. They can continuously update and expand the knowledge base as their product evolves.

💻 Gamma for Web Design Mockups

Here’s a tool that came up: Gamma.

One marketer uses it to mock up web pages before sending them to designers. Instead of writing a brief that says “put the hero here and benefits there,” they create a visual mockup in Gamma.

Their designer gets something concrete to work from instead of trying to interpret a document.

The trick? Use Gamma for design and layout. Not content.

For content, this marketer has a trained ChatGPT that knows their product and PMM best practices. That spits out better copy than Gamma ever could.

But for quickly visualizing how a page should flow? Gamma wins.

📞 Attention: Call Intelligence Gold

Another tool that got people excited: Attention.

It’s call listening software, but smarter.

The problem every marketer knows: getting sales to log things correctly in the CRM. You ask them to note competitors mentioned. To track objections. To capture details.

And then…nothing. Or half-baked notes. Or just silence.

Attention automates all of it.

Someone mentions a competitor on a call? It automatically fills out that CRM field.

But here’s the part that makes it worth it: it integrates with Slack. Creates a bot. The product team can ask “what are the top questions customers are asking this week?” Marketing can pull insights without listening to hours of calls.

👩‍💻 Reloom for Website Wireframing

Moving a site from WordPress to Webflow? Reloom makes it easier.

You describe a page. Reloom suggests sections. Put social proof here. Features here. Testimonials there.

It gives you a starting point instead of a blank canvas.

The workflow: Describe the page in Reloom. It creates sections. Export to Figma. Style it to your brand. Push to Webflow.

One creative lead now owns the entire website. No developer needed for updates.

📝 Clay for Outbound at Scale

Clay came up multiple times. But one marketer shared their full workflow that’s running right now.

They use Clay to build target lists by ICP and geography. Clay agents research each account. Pull information. Fill in details.

That research feeds into personalized landing pages. The landing page builder takes the Clay data and customizes pages for each account.

Clay also creates personalized LinkedIn ads. Pulls screenshots of the target company’s website. Grabs their logo. Generates ad copy based on the research. Creates the actual PNG creative.

The entire outbound process is automated inside Clay. Target list. Research. Ad creatives. Landing pages. All flowing together.

Their pilot results? 500% higher engagement rates compared to generic ads.

The catch? It only works for accounts with 300+ employees because of LinkedIn’s targeting minimums. And there’s always a question of how LinkedIn will react to API-driven campaigns at scale.

But for now? It’s working.

🛠️ The Tools That Keep Coming Up

Standard tools everyone’s still using:

• ChatGPT (often with custom training)
• Claude
• Gemini

But the interesting stuff happens when you combine them with:

• Clay (for data enrichment and workflow automation)
• Zapier (connecting systems)
• Gamma (quick design mockups)
• Reloom (website wireframing)
• Attention (call intelligence)

And for website personalization, tools like Mutiny and Tofu came up. Though most people are still hesitant about IP-based personalization because the accuracy isn’t quite there yet.

But a clever workaround? Instead of dynamically changing content based on IP detection, show a banner at the top. Something like: “Are you visiting from Netflix? Check out this page for you.”

If you’re wrong? No big deal. It’s just a banner they can ignore.

💵 The Cost Reality Nobody Talks About

Clay can get expensive fast if you’re doing heavy enrichment.

But if you already have subscriptions to data tools, integrate them into Clay. Authenticate your existing accounts. Clay won’t charge you credits for using tools you’re already paying for.

Otherwise, running Clay agents across large account lists burns through credits quickly.

💡 What This Means for You

You don’t need a complex AI tool stack.

Most of the marketers we talked to are using 3-4 tools beyond the standard ChatGPT.

They’re not trying to automate everything. They’re solving specific problems:

• Sales needs faster answers? Build a chatbot.
• Designers need better briefs? Use Gamma.
• CRM data is garbage? Use Attention.
• Outbound needs personalization? Use Clay.

Pick one annoying thing in your workflow. One repetitive task. One place where you’re the bottleneck.

Then find the simplest tool that solves it.

You’ll get more mileage from implementing one thing than from researching ten things you never use.