Why old-school sales work still wins in the AI era | Graham Moreno (Head of GTM, Parallel)
Summary
Enterprise AI adoption fails 80% of the time without structured sales and change management—not product superiority. Graham Moreno reveals that old-school onboarding, roadshows, and relationship-building still outperform product-led growth, especially when companies send teams to train customers on workflows rather than letting tools 'spread like wildfire.'
Key Takeaways
- Structured enterprise rollouts with dedicated training and change management show measurable success at 6+ months, while self-serve marketplace approaches deliver minimal change. Moreno saw firsthand that companies throwing multiple tools at employees without guidance achieved poor adoption versus those with step-by-step execution plans.
- Enterprise buyers don't want vendors—they want opinion-driven partners. Customers specifically valued Windsurf sending reps globally to sit with teams, conduct discovery on workflows, propose new processes, and provide ongoing training. This differentiator was rare: no other AI tool vendor was doing on-site change management.
- Change management dictates enterprise success more than technology itself. Every employee pulled into training needs a clear value prop with specific 1-week and 12-week impact metrics. Without this clarity and relationship-building, adoption stalls regardless of product quality.
- Relationship-based selling creates defensibility. Systems integrators maintain multi-decade relationships worth tens of billions because trust and human continuity drive adoption at scale. Replacing this with pure product-led growth in enterprise loses a critical competitive moat.
- AI software founders are overcorrecting against traditional sales playbooks (Mongo, Datadog, Snowflake) with anti-sales sentiment. The reality: 90% of old playbooks still apply in enterprise—the middle ground is reinventing specifically for AI customer expectations, not abandoning sales entirely.
Related topics
Transcript Excerpt
Maybe a place to start that that might be particularly interesting to hear your take on is you've been bringing software to market for a while and you had a huge chapter um call it pre-hat GPT or pre-LM as we know it then post sort of chat GPT in one of the leading companies that's bringing codegen to market now you're joining another company that's at the forefront of this technology. What's like your observations in let's start with kind of new world old world of selling these two different forms of software and do you think it tends to be more different and you're relearning your job or most of it's 90% identical? One of the funny things about this generation of of companies and of sales is that it feels like the playbook companies of like the cloud era, the like mongos, data dog, snowf…