In the span of one week, OpenAI lost two of its most senior sales leaders – both back to Salesforce. That’s not a coincidence. It’s a signal worth paying attention to if you’re thinking about where enterprise Go-to-Market (GTM) motion is actually heading right now.

Kaylin Voss, OpenAI’s VP of sales for the Americas, resigned after just five months in the role – following her former boss out the door, a week after Denise Dresser’s own departure. OpenAI has since brought in Dali Rajic – widely known in enterprise sales circles as the “Croatian Sensation” – as its new Chief Revenue Officer, tasked with shoring up enterprise revenue ahead of a likely IPO. That’s a lot of sales leadership turbulence for a company trying to close large, complex enterprise deals at scale.

The Infrastructure Problem Hiding Behind the Headlines

Here’s what the coverage mostly missed. The Salesforce returns aren’t just about culture fit or comp packages. They point to something more structural: selling AI at the enterprise level is genuinely hard without mature RevOps infrastructure underneath it.

Enterprise buyers are sophisticated. They want defined Ideal Customer Profiles (ICP), repeatable qualification frameworks like MEDDIC, and clean sales pipeline visibility before they’ll commit to six- or seven-figure contracts. Salesforce has spent decades building the tooling and institutional knowledge to support exactly that kind of motion. OpenAI, for all its product momentum, is still assembling those foundations. Experienced sales leaders who’ve lived inside mature CRM ecosystems know what good looks like – and they know when it isn’t there yet.

The TCS-Porsche deal this week reinforces the same point from a different angle. TCS bagged a $1.45 billion engagement with Porsche, acquiring the carmaker’s IT arm for $373 million in the process – its second European mega-deal in eight months. Deals of that size don’t close without serious enterprise sales infrastructure: forecasting discipline, executive alignment, and a sales cycle process that can handle 18 months of complexity without falling apart.

What Developers Are Actually Building Right Now

While the headline drama plays out at the leadership level, something quieter is happening in the tooling layer. The recent releases of fancy-hubspot 0.3.0 and 0.3.1 on PyPI – a lightweight Python library for HubSpot that skips the vendor SDK entirely and works via plain HTTP – reflect a real and growing preference among engineering teams. They don’t want bloated integrations. They want thin, testable, composable connectors they can own and modify without waiting on a vendor’s release cycle.

This matters for RevOps and sales tech teams because it changes where customisation happens. More teams are building their own data pipelines into CRM systems rather than relying on native connectors, which means the CRM itself becomes more of a data destination than an all-in-one workflow tool. If your stack isn’t designed to receive clean external data gracefully, you’ll feel it in your sales forecast accuracy before you feel it anywhere else.

Conference Season Is Shaping Up Around the Same Themes

Digital Summit Philadelphia – running September 23-24 at The Bellevue Hotel – has pulled together a speaker lineup that reflects where marketing and sales tech thinking is converging. Leaders from HubSpot, NASA, the Philadelphia Eagles, and retail subscription brand Nuuly are all on the agenda alongside bestselling author Ann Handley.

That mix is deliberately cross-industry. The Eagles’ marketing operation runs data-driven fan engagement at scale, and it has more in common with a B2B SaaS GTM team than most people expect. NASA’s communication challenges map closely onto the kind of multi-stakeholder enterprise selling that CROs like Rajic deal with daily. The presence of Nuuly – a fashion subscription service – signals that subscription Net Revenue Retention (NRR) thinking is bleeding further into how non-SaaS brands approach customer relationships.

Events like this are worth attending not for the keynotes, but for the hallway conversations about what’s actually working in people’s stacks. A 10-minute chat with a HubSpot practitioner will give you more signal than most vendor webinars.

What CRM and Sales Tech Teams Should Actually Do With This

There are a few concrete takeaways here, and they don’t require you to care about OpenAI’s org chart.

If your sales team can’t articulate a repeatable qualification process, no amount of AI tooling will fix your win rate. Methodology first, automation second.
Lightweight API-first integrations are winning over heavyweight SDK dependencies in engineering teams. When you’re evaluating tools, ask how your developers want to connect them – not just whether a native integration exists.
Large enterprise deals – like TCS-Porsche – close because of process maturity, not product novelty. Your Annual Recurring Revenue (ARR) targets won’t be hit by the AI tool you just licensed; they’ll be hit by the process discipline you build around it.
Talent follows infrastructure. If experienced sales leaders are gravitating back toward mature CRM ecosystems, that’s information about where execution is actually possible right now.

The OpenAI sales story will keep developing as Rajic settles in and the IPO clock ticks. But for most CRM and GTM professionals, the real lesson isn’t about any one company’s drama. It’s that strong sales tech infrastructure is what keeps good people in their seats – and what lets them close deals when they are. Browse the CRM Tools Directory if you’re auditing your own stack, and if you want to stay on top of how this space shifts week to week, the CRM Daily Newsletter is the easiest way to do it.

Why OpenAI’s Sales Exodus Tells You Something About Your Stack

CRM Daily Digest

Get this kind of intelligence every morning.