Conversation
Highlights

How Seismic’s VP of Marketing Eliminated MQLs and Deployed 30+ AI Agents Across Her Team

Most B2B marketing teams measure success by MQL volume. Evelyn Swaim joined Seismic as VP of Global Growth Marketing and eliminated MQL reporting entirely within 18 months.

In a recent episode of Unicorn Marketers, Evelyn shared how she transformed a 65-person marketing organization from chasing top-of-funnel metrics into a pipeline-owning revenue engine. The shift required re-instrumenting systems, changing compensation structures, and fundamentally rewiring how marketing operates alongside sales at the AI-powered enablement platform.

The Hidden Cost of MQL-Driven Marketing

When Evelyn arrived at Seismic, marketing operated like most demand generation teams: optimizing for volume at the top of the funnel. “When I first got here they were focused on top of funnel all around MQLs. They reported on MQLs,” she explains.

The real damage wasn’t the metric itself—it was the behavior it incentivized. Content syndication vendors delivered leads that checked demographic boxes but failed the quality test. “The pay per lead program that you may have had with that house that’s doing the content syndication isn’t going to guarantee you it’s the right Persona at the right title level in your named accounts,” Evelyn notes.

Lead decay compounded the problem. By the time contacts reached SDRs a week after downloading assets, context had evaporated. “That buyer forgot why they downloaded it and they didn’t even realize it was Seismic. They’re thinking it’s from TechTarget or some publishing house,” she explains. The attribution gap meant Seismic paid for leads that sales couldn’t convert—and didn’t trust.

Starting with RevOps, Not Sales

Evelyn’s transformation began with an unconventional partnership: revenue operations, not sales leadership. “Develop a relationship and partnership with your revenue ops organization,” she advises. “The revenue ops team are the ones that’s kind of setting targets.”

Her opening question reframed the entire conversation: “What are our bookings targets? How do we reverse engineer that to understand what we need to bring in? How many deals do we need to bring in? What does the pipeline number need to look like?”

This backwards planning from revenue to pipeline requirements created immediate alignment. RevOps leaders are “very much data driven people,” Evelyn notes. “If you give them the data, if you show them the research of what’s happening in the buying journey today, they’ll get it right away.”

With RevOps bought in, Evelyn had air cover to rebuild dashboards, change lead scoring thresholds, and shift compensation away from activity metrics. The partnership proved more valuable than executive mandate because RevOps controlled the systems and reporting infrastructure required for transformation.

Proving Quality with Velocity Metrics

Sales teams needed proof that abandoning volume for quality would work. Evelyn’s approach: measure conversion velocity by source, not just conversion rate.

“We started to measure velocity and quality, not just pipe, not just volume,” she explains. The team compared how fast tightly-filtered, ICP-fit pipeline moved through stages versus high-volume, loosely-qualified leads.

Her hypothesis to sales: “What if what we bring into the funnel are more sales ready, tightly fit to the ICP in the right Persona with market ready in terms of intent, how fast is that throughput?”

The data validated the shift. Velocity metrics showed that quality pipeline not only converted at higher rates but moved faster—compressing sales cycles and improving forecast accuracy. This evidence created the “aha moment” that made eliminating MQL reporting possible.

Reframing Marketing’s Value for Enterprise Sales

The shift from MQLs required educating sales teams on a reality they experienced but hadn’t articulated: enterprise deals no longer involve single contacts.

Evelyn’s approach was conversational, not prescriptive: “Let’s understand from a selling perspective, from a seller perspective, like what happens when you’re in a deal? Who do you talk to? Who have you had to pull in? What happened when you lost that deal?”

The pattern emerged immediately. “If you’re a sales led enterprise company, you’re going to need to talk to 10 or 12 people to get buy in. That buying committee is getting larger,” she explains.

This insight repositioned marketing’s role entirely. Instead of delivering individual contacts, marketing would help sales map and reach buying committees. “Marketing can help you get in front of the full committee and give you signals on where they are in their buying journey,” Evelyn told sales leaders.

The MQL model optimized for delivering one contact—usually an influencer. The new model optimized for account orchestration across 10-12 stakeholders. Once sales leaders understood this distinction, resistance to killing MQLs evaporated.

Ruthlessly Cutting Spend on Wrong-Persona Sources

With alignment from RevOps and sales, Evelyn systematically eliminated tactics that generated noise instead of signal.

Content syndication went first. Beyond delivering wrong personas, the tactic created a timing problem: “By the time it comes over to your SDRs and your selling teams…might be a week later, and then you have lead decay.” If the team kept syndication at all, “we’re going to give it a very low threshold score. That buyer Persona still has to do many other different things before it pushes over.”

Third-party trade shows followed. “The team was all about scans and MQLs,” Evelyn explains. Badge scanning generated volume but rarely delivered qualified accounts.

The redirected budget went to what she calls ABX—account-based experience—not traditional ABM. “We’ve done a lot of one to ones. We’re doing scaled ABX now with my demand center team, we’re doing a lot of that one to few ABX,” she says. The shift prioritized curated first-party events, partner co-marketing, and targeted programs to named accounts with confirmed buying intent.

The “Jobs I Hate” Framework for AI Deployment

While restructuring demand generation, Evelyn simultaneously deployed AI across her organization. Her team now operates with 30+ AI agents and assistants.

The transformation started with education, not technology. “We actually brought in a consultant, Lisa Adams,” Evelyn shares. “We brought in Lisa and her partner Tani to really kind of give the teams the knowledge or the visibility into what the possible is with AI.”

The first use case targeted a concrete pain point: localization. The team built a custom GPT for translation that eliminated “thousands of dollars per month spent on localization.” More importantly, it collapsed timelines. “People were spending days translating one document and now with the GPT is within minutes,” Evelyn explains.

Her framework for identifying AI opportunities is disarmingly simple: map the jobs to be done for each role, then ask, “What of those jobs to be done do I hate? God, I hate doing that. That’s a time suck for me. Oh my gosh. I keep putting that off because I just don’t like doing this kind of thing. Can an AI do that?”

This created permission for teams to identify automation candidates without technical expertise. Tasks like VLOOKUPs, data normalization, contact scrubbing, and asset versioning to ad specs became immediate targets.

Visualizing AI in the Org Chart

Evelyn formalized AI adoption by requiring it in organizational planning. “Every single one of my leaders, as we’re going through planning right now, I’ve said to them when I want to see in the org chart is their human teammate and then their AI teammate and who’s managing that AI teammate,” she explains.

This visualization forces clarity on what each AI agent handles and who owns maintaining it. An individual contributor might have an AI teammate handling data analysis. A field marketer might have an AI agent managing contact normalization and heat mapping.

To systematize AI expansion, Evelyn created a new role: “I just brought one of my program managers. I’ve just given her. One of her roles now is to be our AI coefficient. And what I want her to think about and help the teams think about is where can we be using AI? When can we have an AI teammate?”

The coefficient role identifies opportunities across functions, particularly in areas where teams assume AI can’t help. For field marketers who “don’t feel like they can rely on AI” because they work with sales on events, Evelyn pushes back: “If you’re thinking about data, instead of doing VLOOKUPs and scrubbing your data from contacts and accounts, give that to AI to do.”

What Marketing Actually Measures Now

Seismic’s marketing organization abandoned MQLs for metrics that correlate with revenue outcomes: “It’s all about pipeline. And did that pipeline move? Did that pipeline progress? Is that converting to bookings? Is that converting to revenue? Did we expand our wallet share within our customers? What’s the value of that pipeline? What’s the velocity of that pipeline? Did we shorten our sales cycles? Did we grow the deal sizes? Did we get higher win rates?”

This measurement shift required complete system re-instrumentation. “If the way that you’re reporting the data is not aligned to the story that you need to tell and how it’s actually impacting the business, you’re going to continue into that rat hole,” Evelyn notes.

She set expectations upfront with executive leadership: the transformation would take 18 months “because there’s behavior change that you have to make. There’s process. There’s the things that we have in our systems. I mean all the things, the instrumentation in your system you have to look at.”

The timeline proved accurate. Changing lead scoring models, rebuilding dashboards to track pipeline progression instead of lead volume, and retraining teams to think in terms of account orchestration rather than contact acquisition required patience and consistent reinforcement.

The result is a marketing organization that signs up for shared revenue ownership alongside sales, measures what drives business outcomes, and leverages AI to operate with efficiency that belies its headcount—all while eliminating the vanity metrics that once defined success.