Marketing’s AI Frontier, Continued: Turning Account Research into Revenue Context

How SupportLogic Built a 90-Second Account Intelligence System That's Transforming Their GTM Motion

Last week, I wrote about how AI is reshaping the workflows of modern GTM teams—from campaign content to post-sale success. I also created a State of AI in B2B Marketing Scorecard that shows what processes marketers are prioritizing and how much impact they are seeing — both BDR/SDR Optimization and Sales Enablement ranked high on the list. Some quiet deployments by top growth marketers in these areas are even more impressive than the flashy demos of vendors. Usable innovation is here!

One of the more thoughtful builds I’ve seen recently comes from Blake Cohlan, Growth Marketing leader at SupportLogic. Blake has been experimenting with AI to streamline execution and improve relevance across the full go-to-market motion—from research and messaging to campaigns and enablement. He’s not building an AI product or chasing hype. He’s tackling a practical challenge every GTM team faces: how to make account context actionable without spending hours buried in transcripts, earnings calls, and fragmented CRM data.

The system he built started as BDR prep. It’s now become something more foundational, providing account research to inform engagement ideas, messaging, and campaigns.

---

The Build: AI-Powered Research, Triggered on Demand

Blake built a system that starts with a click.

From inside Salesforce, any team member can trigger an account research workflow: Run Company Analysis.

That sets off a Zapier-powered sequence that pulls in both external intelligence and internal opportunity data, returning a structured research brief, a historical opportunity analysis, and personalized messaging recommendations built on real-time context. The brief summarizes inputs to provide a synthesis of the following:

• Business Model, Industry Size, and Primary Market

• Key Products or Services

• Leadership Priorities (by executive)

• Market dynamics (Competitors, Positioning, and Partnerships)

• Buying Signals

• Challenges and Opportunities

• Growing Revenue / CSAT / Retention

• Job Listings (Growth Areas, Top Issues, Insights from Open Roles)

• Engagement ideas

• Messaging ideas by persona

Insights are based on and structured around the Account Planning template SupportLogic uses for strategizing sales for target accounts.

The stack includes:

• Perplexity scans public sources with a structured prompt to surface company overviews, executive priorities, market dynamics, funding activity, hiring trends, and competitive signals.

• OpenAI transforms those inputs into structured summaries, past deal recaps, persona-specific messaging, ABM campaign themes, and personalized email openers for identified buying group members.

• SEC.gov APIs bring in added depth by extracting risk factors and strategic priorities from earnings call transcripts and 10k public filings.

• N8N supports more advanced automation where additional logic is needed.

• Zapier orchestrates the entire workflow, managing triggers, API calls, and syncing results back into Salesforce.

The system runs on demand, giving teams a refreshed view of an account when they need it. It’s fast, cost-aware, and doesn’t rely on a constant data stream to be effective.

Early results point to meaningful time and cost savings. A version of this workflow developed in partnership vendor had previously cost around $12 per run. By building it internally using Zapier, the team reduced the cost per run to just $0.30 to $0.80, while gaining full control to refine, extend, and integrate it more deeply over time.

One AE reported saving one to two hours on heavy prospecting days. Research that once took 45 minutes per account is now generated in under 90 seconds. Just as important, the outputs are now structured, repeatable, and consistent, regardless of the researcher. Blake shared this comparison chart:

Process & Challenges

Building the system wasn’t without its bumps. Blake leaned heavily on what some call “vibe coding,” using OpenAI itself to troubleshoot prompts and refine script logic. He quickly learned how easy it is to burn budget while testing models or debugging workflows. He also encountered indexing gaps with Perplexity when key executive signals didn’t surface in the results. That’s led to new experiments combining outputs from multiple LLMs to generate more complete research, rather than relying on a single model.

Developing this process took more than three months as Blake and the team looked at vendors and juggled the project amidst different projects/priorities. The actual vibe coding part of the project, after refining tools and inputs, took just 5 days.

Blake recently shared the full build in a masterclass at the AI x Marketing Summit, alongside Reid Robinson, AI Product Lead at Zapier. He walked through every step, complete with prompts and code, so others could adapt this AI-powered foundation for their own GTM workflows. Here’s a sneak peek at the process:

How One Workflow is Fueling New Experiments

This research foundation has now become the launchpad for a series of focused experiments across the GTM org:

• BDRs are incorporating research insights into their call sequences to increase conversion and relevance in early conversations.

• Automated outbound systems like GemE from UserGems are being used to test message variants pulled from the research brief to validate personalization at scale.

• 1:1 content nurtures are being designed to engage buyers based on their company context, persona needs, and timing signals.

• Social listening is becoming actionable. Signals from platforms like Reddit are now feeding into GTM workflows, helping the team craft timely, contextual responses that meet buyers in their journey and add real value.

• CRM metadata such as campaign engagement, meeting history, and field-level values are being added to make the research layer smarter and more responsive to in-market activity.

These are early tests, but the direction is clear. What began as research enablement is evolving into a responsive GTM system—one that learns from every interaction and adapts how the team shows up.

Key Takeaways

1. Start Small, Scale Fast: Blake's system began as a BDR prep tool but quickly evolved into a foundational research layer supporting the entire GTM motion. Begin with a focused use case, prove value, then expand.

2. Integration Matters More Than Innovation: The real power of this system comes from connecting existing tools (Zapier, Perplexity, OpenAI) with business systems (Salesforce) to create workflows that fit naturally into how teams already work.

3. Experiment with Different Tools for Better Cost-Efficiency: Experimenting across different systems and vendors helped SupportLogic reduce per-run costs from $12 to under $1 while gaining control over customization and integration.

4. Structure Drives Adoption: The system delivers insights in a format aligned with the company's existing Account Planning framework and Salesforce usage making the outputs immediately actionable across teams without additional translation.

5. Measure Time, Not Just Features: The most compelling ROI comes from time saved (45 minutes to 90 seconds per account) and the consistency of output, regardless of who triggers the research. Blake is looking to ultimately measure conversion rate and revenue impacts over time. The system is only as valuable as its ultimate contribution

From Hacks to Systems: The Future of GTM Is Already Being Built

This deployment by Blake is a great example of AI in action - where usefulness wins.

This morning, in a few hours, I’m interviewing Wade Foster, Zapier’s Co-Founder & CEO, on how teams can transform their orgs with AI at the Agentic AI Summit - I’m seeing Zapier have big momentum in the space, with SupportLogic and many other marketers I know. Register to watch the session or the replay here.

I am not sponsored by Zapier - I’ve just been running into them frequently as I’ve researched marketers building impactful real-world AI deployments.

And real-world implementations are indeed here — helping teams move faster, show up smarter, and stay aligned.

If you’re building something similar or testing your own ideas, I’d love to hear it!

---

Carilu Dietrich is a former CMO, most notably the head of marketing who took Atlassian public. She currently advises CEOs and CMOs of high-growth tech companies. Carilu helps leaders operationalize the chaos of scale, see around corners, and improve marketing and company performance.

---