Who Owns the Data? with Keith Jones, GTM Systems Lead at OpenAI

Welcome to Season 3 of 10/10 GTM: The Podcast for Revenue Leaders!

Our guest for Episode 73 is Keith Jones, GTM Systems Lead, OpenAI. Keith is a college dropout turned “professional nerd and technologist.” Prior to joining OpenAI, he held leadership roles at Gartner, MURAL, and Zenput.

In this episode, Ross and Keith discuss how to be a good partner to your Systems team, the smartest ways to evaluate and buy technology, and how to balance the demands of finance with the needs of your customers.

Listen to the episode here, and get the key takeaways from our conversation below.

Being a Good Partner to Your Systems Team

Based on Keith’s experience, there are three core tenants that benefit from a productive partnership between a revenue organization and its GTM Systems team. They are: 

1. Accept some process: To be a good partner to your Systems team, the first step is to make room for planning and testing. “Before any development begins, the Systems team must understand the end goal of what the solution is trying to achieve. That means carving out enough time to plan, test, and validate so the work remains methodical,” Keith explains. “This can be a slippery slope of too much process or process for the sake of process. The key is finding just the right balance — where the process clearly adds value but doesn’t become an end in itself.”

2. Allow time to measure twice, cut once: It takes time to bring new features and functionalities to life, which is why it’s important to give your Systems team the time and space to test and refine. If you rush through development, it often leads to mistakes or missed opportunities. The bottom line is that no one wants to rebuild features that could have been done right from the get go. This means you allow time for thorough development to minimize rework. .

3. Drive alignment and give the team space to work: Another core tenant is to keep priorities clear and consistent. This way, your Systems team will have the space they need to build the right solutions. “There’s a lot of data, insights, and everything else created within these systems, tools, and applications,” Keith says. “It’s important to give us the space to understand and drive alignment. The one thing that is really counterproductive to high-quality systems work is shifting priorities. When you ask us to build something, we can’t do it in an afternoon. We need to figure out how we’re going to build it, actually build it, test it, and then figure out how to get it live. There are a lot of steps in that process. If we’re not aligned at the very beginning or given the space to figure things out, we’ll end up starting and stopping over and over again, never really completing anything good.”

Keith’s main takeaway? Your System’s team wants to partner with you in a way that’s impactful and mutually beneficial.  “Every Systems team worth working with wants to build a lot of really cool things on behalf of their organization,” Keith says. “But they want to do it in a way that won’t require constant fixes, tweaks, or iterations because it was built hastily or haphazardly.”

To make sure you’re establishing a solid partnership, ask questions like: 

• How do you like to work?
• What can we do to make you more efficient, productive, and successful in our collaboration?

This will help you find the right balance and lay a strong foundation for the relationship to evolve toward collective goals. 

Buying Technology

From Keith’s perspective, buying and using technology is a highly collaborative process.  “What we do at OpenAI is put everyone involved in the evaluation of a new type of software through a brief exercise,” he explains. “We ask them to do two things — fill out a doc outlining their five must-haves and up to ten nice-to-haves about this technology. Then, I ask them to take that document, open up a screen recorder, and narrate it for me for five minutes.”

This exercise serves a dual purpose: it prompts each individual to self-edit and clarify what they actually need, while also providing richer context behind those needs. “This is where the AI comes in,” Keith continues. “I take the doc, give it to chat, and say, ‘Synthesize these things together.’ I do this for every individual in the group, giving the chat a bit of detail about each person’s role or persona. What I end up with is a list of synthesized requirements by persona, sourced from a series of individuals within each persona.”

From there, Keith instructs the AI to find overlaps in requirements. “Not everyone is a unique snowflake,” he notes. “We all have distinctive needs, but there’s going to be some overlap. Once I have X number of requirements per persona for Y personas, I want a more concise, consolidated list. Now I have a rubric by persona, and I can show each person how we’ve captured their requirements. We can score every vendor against these requirements, then aggregate those scores into a simpler set of priorities. That makes for a clear, objective decision on where we should spend more time as we move further into the evaluation.”

By combining human input with AI-driven synthesis, Keith’s method keeps everyone engaged, and makes the buying process easier because it reflects the actual needs of the users, versus just the “nice to haves,” and provides a transparent, data-driven basis for selecting the best technology.

Balancing Work for Finance Vs. Customers

To balance work for finance, versus the needs of the customer, Keith employs two strategies: 

1. Reserve a certain percentage of time for quality of life improvements: Keith believes in carving out part of your team’s bandwidth for smaller, “quick win” projects that directly improve day-to-day life for frontline employees. This doesn’t mean sidelining big infrastructure initiatives, but rather acknowledging that incremental tweaks can have an outsized positive impact on productivity and morale.

“One thing that’s really easy to forget, but that I’m reminded of on a regular basis, is that I have a job because this go-to-market team exists,” Keith says. “If I’m not applying some of my bandwidth to making their lives easier and their work more efficient, then I’ve missed the mark entirely. Yes, I have a responsibility to the business to build infrastructure and take big swings, but I’ve got to balance the tradeoff to some degree.”

This might look like dedicating 80% of your time to core projects and 20% to smaller-scale improvements such as cleaning up fields in Salesforce to reduce friction for individual AEs. The exact split will vary, but having a defined ratio encourages you to invest in continuous enhancements that benefit end users.

2. Take a methodical approach to data management: Beyond quick wins, Keith advocates for a systematic way of handling data so you’re meeting finance’s reporting needs and empowering customer-facing teams. Here’s how: 

• Identify critical data points: “Figure out which data is truly essential for running the business,” Keith explains. “What would be an awkward moment in a board meeting if it’s missing or wrong? Not every data point is equally critical. Some are purely nice-to-have. Others are the basis for important decisions.”

• Stack rank priorities: Next, determine which errors or omissions would have the most severe consequences. Keith recommends you ask yourself: “Which data problem would cause ‘anaphylactic shock’ versus which would just cause a little itchiness?” to understand how to rank your priorities.
• Automate wherever possible: Tackle each critical data point from worst to least painful, looking for ways to remove manual burden and reduce human error. “Yes, you might want automation because humans are fallible, but it’s also about empathy — removing the burden from the account director, CSM, or salesperson. They shouldn’t have to spend time inputting data manually if you can capture it automatically.”

By following these steps, you can build reliable, accurate data flows that finance can trust, without overloading your GTM teams with tedious administrative tasks.

About Keith

Keith is a self-proclaimed “professional nerd,” known for his penchant for GIFs and emojis. As the Head of GTM Systems at OpenAI, he leads the efforts behind its GTM Systems function. In his own words, he’s “obsessed with equipping revenue influencers with the best tech teammates that blood, sweat, tears, and cash can buy and configure.”

To learn more about Keith or follow along on his journey, connect with him on LinkedIn.