Building Top-Performing Sales Orgs: Hard-Won Lessons from Scaling 4 Unicorns (Including OpenAI’s Hypergrowth from 10 to 500 People)
Maggie Hott has spent 15 years building go-to-market teams at four unicorns that collectively represent over $500B in market value. She started as the 2nd SDR at Eventbrite, became the first sales hire at Slack (helping scale from $50M to $1B ARR and a $27B Salesforce acquisition), served as Director of Sales at Webflow (scaling from $40M to $140M ARR), and now leads go-to-market at OpenAI where she built ChatGPT Enterprise from scratch. She also runs a venture fund with seven other women investors, backing 30+ founders. These are her personal views, not those of OpenAI.
After 15 years building sales teams at four different unicorns—Eventbrite, Slack, Webflow, and OpenAI—I’ve learned that scaling go-to-market isn’t about following playbooks. It’s about knowing when to throw them out entirely.
At OpenAI, we went from 250 people to nearly 1,000 in a single year while launching what we believe is the fastest-growing enterprise application in history. Here’s the tactical playbook that got us there.
5 Unexpected Learnings That Changed Everything
1. One Exceptional Hire Beats Three Average Hires (And It’s Not Even Close) We held the line on hiring standards even when it slowed us down during hypergrowth. A single mis-hire costs over $1M when you factor in hiring time, ramp time, failed deals, cultural disruption, severance, and starting over. Exceptional people know other exceptional people—your A-players become your best recruiting engine.
2. The “Logo Trap” Kills More Startups Than Bad Product-Market Fit Don’t confuse big company brands with startup readiness. I see this constantly—founders hire someone from Google or Salesforce assuming they can build and scale. But you need to understand: what did they actually build? What did they scale from scratch? You need builders, not operators of existing machines.
3. Frontline Managers Are Untrained Therapists (And That’s Your Biggest Risk) Your frontline managers shape your team’s daily experience. They’re your organization’s pilots—and you wouldn’t board a flight with an untrained pilot. Yet most companies underinvest in manager development. These people are shouldering emotional weight while managing performance. Overinvest in their leadership development, executive coaching, and emotional intelligence training.
4. Customer Losses Channels Matter More Than Customer Wins Channels Everyone loves the feel-good wins channel. But the losses channel is where magic happens. At Slack, WebFlow, and OpenAI, as soon as that channel lights up, everyone swarms in—our COO, engineering leaders, product managers. They want to know why we lost and what we could have done better. Build this from day one.
5. Most Decisions Are Two-Way Doors (Stop Treating Them Like One-Way) Amazon’s door model changed how we think about change management. One-way doors (like major pricing changes) are hard to reverse. But most things—reorganizations, pilot programs, new verticals—are two-way doors. If you think it’s right, move fast. Don’t wait for the “perfect” timing or end of fiscal year. In AI, speed beats perfection.
Building ChatGPT Enterprise from Zero: The Inside Story
In early 2023, OpenAI had just launched the fastest-growing consumer app in history. But we had a hypothesis: ChatGPT Enterprise would need a completely different go-to-market motion than our API business.
When I joined, our entire sales and go-to-market team was less than 10 people. No SDRs, no SCs, no CSMs, no sales ops, RevOps, marketing enablement—not even a working Salesforce instance. Just six account directors and one technical success partner, all focused on API sales.
We made a deliberate choice: build a dedicated go-to-market org from scratch rather than trying to retrofit the existing team. Here’s exactly how we did it.
The Build vs. Adapt Decision
Most companies would have tried to enable the existing team to sell both products. We went the opposite direction and created a “tiger team” dedicated solely to ChatGPT Enterprise. Why?
- Speed: No competing priorities or split focus
- Customer clarity: Enterprise buyers needed specialized expertise
- Product complexity: ChatGPT Enterprise required fundamentally different conversations than API integrations
This approach gave us speed, clarity, and deep customer focus during a critical window. The result? What we believe launched as the fastest-growing enterprise application in history.
The 0-to-500 Scaling Framework
Phase 1: Foundation (Months 1-3)
- Hired 3 enterprise AEs with specific vertical expertise (financial services, tech, healthcare)
- Built basic sales processes and qualification frameworks
- Created initial pricing and packaging structure
- Established customer success foundations
Phase 2: Systems (Months 4-9)
- Added SDR team and inside sales capability
- Implemented proper CRM and sales operations
- Built enablement programs and competitive intelligence
- Added specialized roles (solutions engineers, customer success managers)
Phase 3: Scale (Months 10-18)
- Hired vertical-specific sales teams
- Added channel partnerships
- Built marketing engine and lead generation
- Created enterprise customer advisory boards
Phase 4: Integration (Months 19-24)
- Unified with API sales team (500 people, new roles, new workflows)
- Cross-trained everyone on both products
- Streamlined processes and eliminated duplication
- Created unified customer experience
The Big Integration Bet
Earlier this year, we made another bold move: we collapsed both orgs. We took 500 people—most who’d been at the company just months—and brought them into one unified go-to-market organization.
Everyone got new roles, new workflows, new managers. Every person had to learn the opposite product they weren’t selling. API sellers learned ChatGPT Enterprise, ChatGPT Enterprise sellers learned API.
This was painful but drove faster execution, eliminated duplication, and most importantly, created a better customer experience. Many customers wanted both products—now they had one unified team to work with.
The lesson: Old playbooks don’t always apply in AI. Design around your products and customers, not legacy org structures.
The Hiring Playbook: How to Build Your A-Team
Hire for Mission Alignment, Not Just Excitement
A mission alone isn’t enough. You need people who can clearly articulate how their specific skills will move you closer to your objectives. At OpenAI, we look for candidates who understand how their work contributes to creating AGI that benefits all humanity—not just people excited by the vision.
Think Teams, Not Just Talent
You can have 50 top performers, but if they can’t collaborate, you’ll bottleneck. Think of hiring like assembling a puzzle—you need complementary pieces that fit together. Skills, roles, and culture all need to align because trust and collaboration matter as much as individual excellence.
The Chaos Translator Profile
I love hiring “chaos translators”—people who thrive in ambiguity. Throw them on a renewal, an outage, or a crisis, and they handle it with a smile. These are your builders. Early on, prioritize versatile generalists over specialists, and avoid lone wolves (there’s a time and place for them, but not in your first hires).
Interview Framework: Tactical + Behavioral
Tactical Questions (Skills & Experience):
- What did you build from scratch?
- What processes did you invent?
- How did you mentor and train others?
- How did you help hire and scale the team?
Behavioral Questions (Mindset & Character):
- Describe a painful deal you lost. What did you learn?
- Who were your closest cross-functional colleagues?
- Tell me about a time you had to pivot strategy mid-quarter.
Red Flags That Mean Immediate No
- Blaming others for failures: Look for ownership and accountability
- Frequent job hopping without progression: You want growth within companies, not just title changes
- Primarily motivated by titles or money: They’ll jump ship for $20K more
- Lack of humility: We all make mistakes—find people who admit and learn from them
Building Teams That Thrive: The Slack Security Breach Lesson
In March 2015, I walked into Slack as their first sales hire. Day four: major security breach. Stuart Butterfield stood on a desk (we were one small room) and said, “This will be challenging, but we’ll make it through with hard work from every single one of you.”
The trust and respect employees had in Stuart was palpable. That’s when I learned: alignment isn’t tested in good times—it’s revealed in hard ones.
Three Pillars of High-Performance Culture
1. Encourage Real Debates Make sure everyone feels comfortable speaking up, from SVP of Marketing to the newest SDR. No “meeting after the meeting.” Once you decide, move forward (you can always revisit with intention).
2. Make Priorities Obvious Leaders, this is on you. Use async, transparent communication to connect daily work to goals. Our CPO at OpenAI sends a company-wide post every Monday morning outlining what’s launching, key priorities, and what to be aware of.
3. Speed Up Decisions Empower decision-making at all levels. In the last year, AEs came to me wanting to test new verticals, pricing strategies, and pilot motions. We experimented with all three. All were wildly successful and changed how we sell today. The people closest to the business have the best ideas.
Customer Obsession as Strategic Advantage
At Slack, we assigned non-sales DRIs (Directly Responsible Individuals) to our largest customers. Our VP of Engineering owned VMware. Our CTO owned Nike. Suddenly, customer feedback hit our roadmap much faster.
Build customer wins AND losses channels immediately. The wins channel feels good, but losses channels drive learning and improvement.
Scaling Without Breaking: The Art of Giving Away Your Legos
When Stuart brought in Bob Fry as SVP of Sales at Slack, many of us raised eyebrows. Why did a thriving startup need a corporate executive? We were completely wrong.
Bob brought structure, scalable systems, and season judgment from scaling Salesforce. He mentored internal talent instead of replacing us. Result: Slack grew from $50M to $1B ARR, went public, and had a $27B acquisition.
The “Give Away Your Legos” Framework
Coined by Molly Graham, this means handing off parts of your job as you scale. At OpenAI, Sam brought in a new CPO (Kevin), CEO of Applications (Fiji), and CFO (Sarah Frier). This unlocks bigger impact—you’re scaling influence, not losing control.
Two Rules for Giving Away Legos:
- Actually give them away (don’t loan and micromanage)
- Start with what drains you most (if sales is working but draining you, hire a head of sales and focus elsewhere)
Blending External and Internal Talent
Your internal people know where the bones are buried. Don’t lose your early believers.
- Internal promotions create cultural continuity and loyalty
- External hires bring fresh perspective and experience
- Be transparent about layering decisions—share the why
- Prioritize clear career paths—growth drives retention
I stayed at Slack for years post-vesting because I saw a strong future ahead.
The Compensation Strategy Nobody Talks About
Companies rarely use compensation as a strategic lever. Here are my hot takes:
Put Everyone on Bonus Plans
Not just sales—everyone. At Slack, we tied every employee’s bonus to ARR and customer adoption. Suddenly, when product shipped EKM (Enterprise Key Management), they saw direct revenue impact. Features shipped faster when everyone’s aligned to the same goals.
Start Sales Teams on Full Salary
We did this at Slack, WebFlow, and OpenAI. Early sales teams are effectively user researchers—you want them focused on building and scaling, not their paycheck. It emphasizes strong foundations over short-term commission chasing.
When you do introduce commissions, be intentional. At Slack, we wanted new logo adoption, so we weighted 75% of comp toward new logos vs. upsells. Behavior followed incentives.
The Go-to-Market Mistakes Rapid-Fire Playbook
Mistake #1: Starting with Enterprise Too Early
When founders tell me they want to go after Walmart or Nike, I push back 10/10 times. Enterprises are tempting—big logos, big ACVs, great for pitch decks—but they’re wrong for early-stage companies because:
- Complex hierarchies and lengthy cycles (rarely under 6 months)
- Deep customization needs that derail your roadmap
- Risk-averse buyers who won’t be first adopters
Instead: Define your ICP by analyzing existing thriving customers. Understand why they bought you and why they chose you over competition. Create target lists of similar companies and go outbound.
Mistake #2: Outsourcing Product-Market Fit
You cannot outsource finding product-market fit. Warm intros don’t scale—you must sell to strangers. As a data point: I’ve had three angel investments fail, all amazing companies, but they couldn’t find go-to-market fit fast enough.
Be the 5%: In the age of AI, 95% of outbound is spam. Be personal, relevant, and helpful.
Mistake #3: Treating Pilots Like Freebies
Last year, we built a dedicated pilot team and treated pilots like real sales cycles. Deal sizes went up 5x. But pilots are resource-intensive, so follow three rules:
- Secure explicit customer buy-in upfront (they have budget to convert)
- Define clear pilot roles and responsibilities (customer must lean in too)
- Executive alignment is mandatory (if execs don’t know about the pilot, the deal won’t close)
Mistake #4: Thinking You Don’t Need Customer Success
“We don’t need customer success” is shortsighted thinking. A deal isn’t won when signed—it’s won when deployed with a happy customer singing your praises.
It costs 5x more to acquire new customers than retain existing ones. Churn isn’t just lost revenue—it’s brand damage. Remember Slack’s Uber crash? We spent three years convincing enterprises we were enterprise-ready after that incident.
The Product-Led Growth Trap
Product founders often forget about sales upgrade paths. PLG drives early adoption and virality, but you need sales teams for upmarket strategic deals. Plan for both from the beginning.
We didn’t do this well at WebFlow. After 7 years of killer self-serve product, when we hired a sales team, there wasn’t enough product differentiation to sell. It took 1-2 years to build enterprise features—that’s wasted time and opportunity.
Design with sales in mind from day one.
The Power of Strategic Storytelling
Nobody wants to be first. Prospects want proof, not promises. Capture customer stories early—get logo rights, case studies, and quantified outcomes.
When OpenAI announced our Moderna partnership (helping reduce FDA approval time for cancer drugs), it unlocked credibility across life sciences. Our Morgan Stanley partnership (wealth advisors with happier customers) opened financial services.
Internally, engineers and product managers love hearing how people use their work. Externally, customer stories build credibility and momentum.
The Four-Year Uber Redemption Story
Remember that Slack-Uber crash I mentioned? Four years later, we won Uber back. It took time, but we earned it when the product matured, infrastructure could handle scale, and timing aligned for both sides.
The right customer at the wrong time is still the wrong fit.
The 4 Biggest Mistakes Maggie Made (So You Don’t Have To)
1. Waiting Too Long to Hire External Leadership at WebFlow I believed we could figure it all out internally. We had incredible people but lacked the external perspective that comes from “seeing the movie before.” When we finally brought in seasoned executives, the acceleration was immediate. Don’t let pride or fear of change delay bringing in game-changing leadership.
2. Underestimating the Enterprise Product Gap at WebFlow We built an incredible self-serve product for 7 years but didn’t design enterprise differentiation from day one. When we hired a sales team, we spent 1-2 years catching up on enterprise features we should have been building in parallel. The result: massive opportunity cost and delayed upmarket expansion.
3. Not Documenting Customer Success Patterns Early Enough At my first few companies, we were great at celebrating wins but terrible at systematically capturing why customers succeeded. We lost countless stories, case studies, and proof points that could have accelerated our sales cycles. Start documenting success patterns from your first 10 customers—you’ll need those stories to convince customers 100-1000.
4. Being Too Nice About Performance Issues Early in my career, I would coach struggling team members for months, hoping they’d turn around. The kindest thing you can do for everyone—the person, the team, and the company—is to move quickly on performance issues. Every day you delay compounds the problem and often makes the eventual conversation harder, not easier.
Building world-class go-to-market teams isn’t about following someone else’s playbook—it’s about understanding principles, adapting to your context, and moving fast when you know you’re right. The companies that win in AI will be the ones that can hire exceptional talent, maintain high performance culture, and scale without breaking their foundation.
