Welcome back to Funnel Lab Fridays, the weekly LinkedIn live session that covers the latest trends and challenges faced by data-savvy marketers today. I’m your host, Eric Westerkamp, CEO of CaliberMind.
Today, I’m joined by Doug Bell, CMO at CaliberMind, and Jordan Crawford, Founder of Blueprint GTM, to discuss how AI-powered sales engagement platforms, like Clay, are helping organizations reach new levels of nuance and sophistication in their marketing and sales outreach.
Meet Jordan Crawford
Jordan Crawford has been knee-deep in B2B growth since 2016. His company, Blueprint GTM, helps people answer any question about their market with bespoke data sets. Rather than manually finding out details about your prospects at low scale, Blueprint can help you answer questions at high scale by sorting your market by companies that have problems you can solve right now. You’ll also get the tools you need to contact prospects in the right channels.
The Changing Landscape of Sales Engagement Platforms
The current landscape for sales engagement platforms can be broken up into waves. You can think of ZoomInfo as a first-wave provider and tools like Apollo as second-wave. These waves are about consolidation or as Jordan calls it, “One tool to rule them all.” The only problem is that these tools are missing something very important: the data set.
While Apollo’s data is good, it’s not nuanced enough to give marketers exactly what they need to differentiate their messaging. You can identify which organizations just raised a round of funding, but you don’t have enough detail to personalize your outreach to their needs.
Third-wave sales engagement platforms, like Clay, do something different, they allow you to invent a dataset. You can chain together analyses, cover gaps in data, and even live research online. Rather than relying on a database, you can create bespoke data at the list level, even at the Total Addressable Market (TAM) level, by deploying AI agents to find that information on the public web.
Bridging the Gap: How Marketers Fit In
The question now becomes, how does that play in for marketers? It’s clear how these platforms can help sales teams automate and orchestrate activities, but how can these tools tie into the movement toward marketing decision engines?
Doug explains that tools like Clay help translate data into action in a meaningful and quick way. They can help you go from concept to activating a prospect or outbound funnel really quickly. You can constantly test, iterate, and improve.
Jordan elaborates saying, “One of the really interesting things I’m doing for a customer is leveraging intent signals. The previous way people did this was with arbitrary scores—half a point here, 1.2 points there.” Now, models like Claude Haiku, which is a competitor to GPT-3.5 at a tenth of the cost, are exceptional. You can feed the models massive amounts of data and say, ‘Given all this information, give me a score on this account.’ And because it’s cheap enough, you can do that every month.”
This process not only helps you identify ideal accounts based on specific context but it can also help you find net new contacts for specific roles. Using tools like Clay, you can close the gap for your entire market using Clay’s models and deployments.
Data to Action: Speed and Efficiency
Doug emphasized the speed and efficiency these tools bring. “With tools like Clay, you can start a buyer journey and see engagement data flowing through your marketing decision engines. You can rapidly adjust, creating a feedback loop that’s unprecedented in my career. You’re going from activation to measurement to change in weeks, as opposed to months or even a year.”
“You don’t have to string this stuff together on your own,” says Jordan, “Clay has 50 different data providers.” Blueprint GTM sees a 90% match on mobile phones and personal emails. You can programmatically deploy to ad channels, use tools like ServiceBell for parallel dialing, and feed into your outbound emails. Suddenly, you have an orchestration layer that can do what humans do, but programmatically.
Navigating Legal and Privacy Concerns
At CaliberMind, we’ve been deploying AI inside our platform. One of the biggest pushbacks we get from customers, especially enterprise organizations, is in regard to data privacy and legal issues. Jordan explains that one of the misunderstandings about AI is that everyone’s data is used for training. “If you pass data to GPT through APIs, the models aren’t training on any API calls,” he clarifies.
While lawyers are getting involved, the reality is that people are building their own point solutions and their own models for specific use cases, many of which will require data privacy, like in healthcare. AI innovation is going to happen.
The Future of Large Language Models and Bespoke Solutions
Jordan explains that “OpenAI has said they have no moat. There are a lot of small language models that come really close to large language models, always on their heels. I think what will happen is you’ll use large language models to explore possibilities at low scale.”
For organizations looking to unearth data insights at high scale, they’ll need to move away from large language models and run their own bespoke models locally, using their own data. Large language models will always run up against certain workflows that can be done better by bespoke training sets. There’s a possibility that new point solutions will arise to fulfill this need.
Why This Represents a Sea Change in Sales Engagement
The way that sales engagement has been done so far is inefficient. Even companies like ZoomInfo ramp up sales teams instead of leveraging their own data. Clay and other similar tools are different because large language models can already speak to buyers like a 12-year-old, and it’s only improving.
“Most personalization today is superficial,” says Jordan. “Hey Doug, you wear shirts, I wear shirts, here’s a bunch of stuff about my product.” The sea change here is that AI can do what your best rep already does and replicate it at scale. Clay is kind of the first tool to let you do that and it took them years to build this flow. You can gather data and deploy it programmatically.
Real-World Example: Programmatic Personalization
Jordan shared an example of Clay in action. He had a customer whose best play was contacting the chatbot on the company page, asking for support, emailing support, and calling support. They’d gather up those answers and send them to the head of customer support, saying, ‘Three reps gave me three different answers on three channels, and they were all different. Is this what you expect of your customer support department?’ That’s the kind of bespoke data play you can run with Clay.
Tool Time: Unveiling the Magic Behind Clay
In our Tool Time segment, Jordan gave us a behind-the-scenes look at how he uses Clay. For one customer, he programmatically found a customer’s competitors by searching Google. Then he structured that information into a competitor list. He built a Technographics API that allows his team to figure out which buyers are using his customer’s competitors.
Jordan also showcased how AI models like Claude Haiku can be used for programmatic emails at a fraction of the cost. “You can allow it to be extra creative,” he said. “Clay will give it the context, and you constrain the model to do specific things. For example, we pulled negative reviews about our customer’s competitor and crafted emails highlighting those pain points without pitching the customer’s value prop at all.”
Closing Thoughts
Thanks for joining us on this episode of Funnel Lab Fridays! We hope you found our deep dive into AI-centered sales engagement platforms as enlightening as we did. The future of B2B marketing is unfolding before our eyes, and we can’t wait to see where it takes us.
