Jeannie shouldn’t wait on gen AI—but she also shouldn’t deploy a fully autonomous, customer-facing salesbot tomorrow. Instead, she should green-light AI now in tightly scoped, human-in-the-loop ways, and treat full customer-facing automation as a capability they grow into, not a switch they flip.
Right now, Jeannie is at risk of what Jim Lecinski calls “act-now, think-later” leadership: chasing FOMO on AI instead of starting from PulsePoint’s real problem—thin margins compared with peers. The question isn’t “Chatbot: yes or no?” It’s “Where can AI materially improve our economics without eroding the very relationship-based service Tyrell and other clients value?”
The case itself suggests a smarter path. Lecinski recommends using AI internally first: prospect ranking, RFP drafting, FAQ retrieval, sales training. That directly boosts sales productivity while preserving the human interface that Mark believes drives upsell and cross-sell. Dharmesh Shah’s perspective pushes the other way—move decisively, because capabilities are improving every six to nine months—but even he advocates framing AI as augmentation, not head-count reduction, and giving customers opt-outs.
So I’d argue PulsePoint should:
- Deploy gen AI inside the sales and service teams first: AI that drafts emails, prepares talk tracks, surfaces past interactions in real time. This tests the tech on lower-risk surfaces but still generates learning and margin improvement.
- Pilot a hybrid chatbot with safety rails. Start with narrow, well-defined flows (simple support questions, scheduling, basic pre-sales qualification) and keep a “transfer to human” button prominent—especially given the case data that 90% of people want that option even if they’re fine with bots for speed.
- Co-design the rollout with skeptics. Mark, Linda, and even Tyrell shouldn’t just be “managed”; they should be treated as design partners.
The real competitive edge isn’t having a salesbot; it’s building an organization that learns faster from using one. PulsePoint should move now—but the product is as much the org change as the chatbot itself.
