OpenAI's rapid ascent in artificial intelligence has been matched by the swift development of its global customer success (CS) organization. In just 18 months, Vanessa Gutiehi, OpenAI's Head of AI Deployment & Adoption, transformed from the company's first CS hire into the leader of a worldwide team. Spanning key cities from San Francisco to Seoul, her mission is to ensure OpenAI's enterprise customers not only derive transformative value from AI but also achieve sustained growth and retention in this era of unprecedented technological advancement.

Top 5 Learnings from OpenAI's AI Customer Success Playbook:

  1. Every team member can be their own biz-ops analyst — Integrating AI with data empowers team members to gain immediate insights, eliminating the two-week wait times for analysis. Customer Success Managers (CSMs) can now proactively identify areas for new assets and enablement, a stark contrast to traditional request-and-wait models.
  2. Map your customer journey to find AI opportunities — A single cross-functional whiteboarding session focused on identifying the top three customer journey friction points can unlock immense value. AI can then automate repetitive tasks, boost CSM productivity, and enable personalized experiences at scale.
  3. Build prompt libraries, not playbooks — Traditional playbooks often fall short. Today's operational rigor comes from developing high-quality prompt libraries and packaged GPTs. These tools ensure consistency and personalization across the entire customer journey.
  4. Race to ROI, not feature adoption — Customers seek outcomes, not just AI tools. The focus should shift from measuring AI feature usage to quantifying the tangible business results those features deliver. While early AI adoption metrics can serve as a "Trojan horse" to capture executive attention, the ultimate pivot must be to true business value.
  5. The best tiger team members aren't your most technical people — Initially, OpenAI tapped its most technical staff for piloting new AI tools, which proved to be a misstep. A more effective approach is to select individuals with medium technical proficiency who excel at communication and relationship-building. Their ability to foster trust and empathy is crucial for driving widespread AI adoption across teams.

Why Customer Success Matters More Than Ever in the Age of AI

Vanessa Gutiehi underscores a fundamental truth for all SaaS leaders: customer success directly fuels business prosperity through increased purchases, longer retention, more referrals, and reduced service costs. However, the advent of AI has significantly raised the bar.

"ChatGPT has changed expectations," noted Lowe's SVP of Digital. "Today's customers demand more value at greater speed and extremely personalized."

This isn't mere theory. Lowe's, for instance, developed AI-powered tools on OpenAI's platform, offering customers 24/7 personalized expertise for home projects. Much of this traffic occurs overnight when physical stores are closed, resulting in measurable sales increases and repeat customers who trust Lowe's with their significant home improvement investments.

For SaaS leaders, the message is clear: prioritize building solutions that provide instant answers, automate support, and accelerate time to ROI.

The Four-Point Blueprint for AI-Powered Customer Success

To avoid what Lowe's CIO Samantha Nye Godable terms "the AI pilot doom loop," Vanessa Gutiehi shared a practical four-point framework:

1. Support Automation (The Low-Hanging Fruit)

Automating at least some tier-one support tickets is no longer optional. The data is compelling:

  • Klarna leverages OpenAI to triage over two-thirds of its tickets, slashing average response times from 11 minutes to 2 minutes and saving $40 million annually.
  • Major enterprises like T-Mobile and leading banks are implementing similar solutions.

A key insight here is that the same agentic technology used for answering tickets can also function internally as a co-pilot for go-to-market teams. The strategy isn't agents versus co-pilots; it's a synergistic approach, depending on the use case and stage of implementation.

2. Onboarding Velocity (The Biggest Opportunity)

Vanessa's tactical recommendation is to run a customer journey AI exercise. Here's the playbook:

  1. Map out three journeys: end-user, admin, and account-level customer journeys.
  2. Identify friction points for both customers and internal go-to-market teams.
  3. Prioritize the top three friction areas for immediate investment.
  4. Project time-to-value reduction resulting from these investments.
  5. Move quickly: build, measure, and iterate.

ChatGPT can be invaluable throughout this process. Even a photo of a whiteboard session with messy handwriting can be translated by ChatGPT into organized action items, eliminating lost notes or delays in documentation.

At OpenAI, the CS team focuses on three core objectives:

  • Increasing CSM productivity through co-pilots and tools that expand capacity.
  • Automating repetitive tasks such as data pulls, customer settings, and mundane administrative work.
  • Adding personalization at scale to create "wow" experiences that center the customer.

The results are significant: OpenAI's CSMs can generate highly customized launch plans in seconds, integrating data from sources like Gong, recent earnings calls, and customer-specific information. Tasks that once took hours are now completed with just a few clicks.

3. Data-Driven Decision Making (The Game Changer)

"This for me has been such a game changer in the way that I lead my business," Vanessa emphasized.

When ChatGPT is integrated with internal data, every team member effectively becomes their own business operations analyst. Vanessa recalled her five years at Slack, where analysis requests would take weeks. Now, she performs analyses herself instantly. More powerfully, individual team members are proactively bringing insights to her, guiding where the business should focus its resources and enablement efforts.

"It's so powerful when now every single person can have these insights and help shape the company in a way that is putting our resources in the most impactful areas."

4. Voice of Customer Programs (The Easy Win)

Modern AI tools make it easy to establish a robust Voice of Customer (VoC) program. By integrating Gong calls and data from community platforms, AI can synthesize patterns and insights. OpenAI is currently heavily optimizing its VoC program, recognizing it as a rich source of data for accelerating product development and fostering customer-driven innovation.

The OpenAI Iteration Loop

OpenAI's competitive advantage stems from its rapid iteration speed. Their model follows this loop:

Research (the "beating brain") → Applied team builds products → Go-to-market deploys to customers → Learn and feed back to research

Vanessa recommends mimicking this loop for AI adoption within any organization:

  1. Ideate (You're your own research lab): Conduct monthly team meetings to pinpoint friction points in the customer journey.
  2. Pilot (Small tiger team): Test solutions with a small group of CSMs across different segments, ensuring they are close to real customer questions.
  3. Measure (Track time savings and new capabilities): Quantify what new tasks or efficiencies are now possible.
  4. Iterate and scale: Enable the wider team on the new tool, celebrate learnings, and roll out broadly.

Critical insight on tiger teams: Avoid selecting only the most technical person. Instead, choose someone with medium technical skills who enjoys communication and building relationships. They are instrumental in fostering the trust and empathy necessary for successful AI adoption across the team.

Key Guardrails as You Scale AI Adoption:

  • Keep humans in the loop, especially during initial stages.
  • Secure buy-in from all necessary partner teams.
  • Choose excellent champions (as described above).
  • Avoid doing 1,000 pilots that go nowhere — maintain sharp focus.

Racing to ROI: Focus on Outcomes, Not Features

Defining customer value is complex, and the proliferation of AI products with diverse pricing models (consumption, token-based, etc.) has made it even more challenging.

The fundamental truth remains: Your customers and employees don't want AI tools; they want outcomes.

Stop measuring whether customers used your AI feature or logged in X times. Instead, measure the concrete business outcomes that AI feature drives. Even when difficult, maintain a laser focus on true business value.

Pro tip: Executives are currently keen on discussing AI feature adoption, as it's still a novel and exciting topic. If this serves as your "Trojan horse"—either internally or with customers—to engage a CEO and then pivot to discussing genuine business outcomes, leverage it. OpenAI has found success with this strategy.

The OpenAI Value Journey for ChatGPT Enterprise

OpenAI's value journey illustrates how to achieve quick ROI:

Phase 1 (Immediate): Measure increases in tool usage, understanding, and literacy through surveys and adoption metrics.

Phase 2 (Short-term): Focus on time savings, productivity gains, and AI fluency.

Phase 3 (Long-term): Aim for transformative business change.

The key insight is that Phase 3 takes time. By demonstrating immediate ROI in Phase 1, you gain the necessary time to work towards the truly transformative business changes that hold lasting significance.

Real example: The San Antonio Spurs launched multiple GPTs focused on customer retention. OpenAI measured AI fluency, literacy, and time savings while simultaneously working towards the ultimate metric: customer retention uplift. The Spurs have detailed their seven-step enterprise AI journey manifesto on their blog.

Adapting to Rapid AI Progress: Change Management That Works

The current pace of AI development is unprecedented. As leaders, it's crucial to establish flexible processes that facilitate continuous adoption of this technology.

The good news is that the change management playbook isn't fundamentally different from past strategies:

  • Secure strong executive sponsorship.
  • Clearly communicate the "why" and the vision (e.g., how will time savings be utilized, how will customer interactions improve?).
  • Provide opportunities for experimentation and play with tools, perhaps through off-sites or hackathons.
  • Reward experimentation and learning.

Critical insight: Build AI automation not just for customer outcomes, but also to free up your team's time to learn new AI features. The landscape will continue to evolve rapidly, and your team needs the capacity to stay enabled on new advancements.

How Vanessa Personally Stays on Top of AI

There's no magic solution; it requires dedicated effort. Here are Vanessa's tactical tips:

  1. Block time to use the tools — Get hands-on with different models to understand their current power and limitations.
  2. Use voice mode religiously — During commutes, chores, or other tasks, leverage voice mode as a personalized tutor to explain complex technical papers, such as new inference time scaling research.
  3. Listen to quality AI podcasts — Vanessa recommends specific podcasts to stay current with the latest developments.
  4. Use ChatGPT Projects — For ChatGPT Enterprise users, Projects can significantly amplify work efficiency. OpenAI's free Academy offers videos on business workflows and use cases.

Her Final Message: Start Today

"This will not slow down, and each of you really needs to use AI as a competitive advantage. How can ChatGPT or other AI tools help you ideate, pilot, and scale for your customers? You need to do it now."

Vanessa's own first-day story at OpenAI exemplifies this urgency: she launched two customers on her second day. While those initial launches "probably weren't the best," by Friday, through continuous iteration, learning, questioning, and leveraging ChatGPT to understand customer businesses, her performance had dramatically improved.

Speed matters. Learning quickly matters. And using AI to empower both customers and teams is rapidly becoming an essential skill.

Top 5 Mistakes Vanessa Made Building CS at OpenAI

During her talk and Q&A, Vanessa was refreshingly candid about the missteps encountered along the way. Here are the key mistakes she and the OpenAI team identified:

1. Picking Only the Most Technical People for Tiger Teams

Early on, OpenAI selected its most technical team members to pilot new AI tools, which proved to be a mistake. A more effective strategy is to choose individuals who are moderately technical but possess strong communication and relationship-building skills. These individuals are crucial for fostering the trust and empathy needed to drive widespread AI adoption across teams. Technical prowess, in this context, was less critical than interpersonal skills.

2. Running Too Many Pilots That Went Nowhere

The "AI pilot doom loop" is a real phenomenon, and OpenAI was not immune. They initiated numerous pilots without clear focus or a firm commitment to scale. The lesson learned was the importance of ruthless prioritization. It's better to select the top three friction points, commit to them, and see them through to scaled deployment, rather than dispersing resources across dozens of unfocused experiments.

3. Over-Indexing on AI Feature Usage Instead of Outcomes

Initially, OpenAI measured AI feature usage through metrics like logins and activation rates. However, customers are not concerned with merely using features; they care about tangible outcomes. The team had to shift its focus from "are they using it?" to "what business results are they seeing?" While measuring business outcomes is more challenging, it is the only metric that truly matters for customer retention.

4. Not Building Enough Personalization into Early Customer Experiences

OpenAI's initial customer onboarding experiences were generic. Vanessa admitted her day-two launches "probably weren't the best." The turning point came with leveraging AI to create hyper-personalized experiences, including custom demos tailored for specific customer teams, training adapted to their exact use cases, and dummy data relevant to their industry. These efforts created "wow" moments that forged genuine partnerships. The mistake was not prioritizing this level of personalization from the outset.

5. Underestimating the Time Needed for Team Enablement

As new AI features and models rapidly shipped, OpenAI initially failed to allocate sufficient time for its go-to-market teams to learn and experiment. The team's intense focus on customer outcomes inadvertently neglected the internal capacity needed for continuous learning. The solution involved intentionally automating work to free up team time for hackathons, experimentation, and dedicated enablement sessions. A team cannot effectively help customers adopt AI if its members are not continuously updated on new advancements themselves.