May 13, 2026

Future-Proof Your Career | AI Upskilling Guide | Justin Trombold | 388

Most companies aren't failing at AI because of bad tools — they're failing because they skip the fundamentals. In this episode, host Jeff Mains sits down with Justin Trombold, President of Antison Advisors and former consultant at Deloitte and Grant Thornton, to unpack why so many AI initiatives stall in the experimentation phase and never create real business value.

Justin brings a rare perspective — rooted in academic research and first principles thinking — to one of the most pressing challenges in business today: turning AI curiosity into measurable results. From diagnosing organizational readiness to rethinking how SaaS providers serve customers, Justin delivers a clear, grounded framework for leaders who want to move from pilots to scaled impact.

Key Takeaways

[0:00] — Intro: Why AI initiatives look impressive but fail to move the business forward

[3:27] — Justin's journey from academia to consulting and how first principles thinking shaped his AI advisory approach

[8:14] — First principles vs. layering AI on top: Start with “what are we trying to solve?” not “what's the newest tool?”

[9:30] — The difference between process-level AI improvement and customer-outcome-level reimagination

[13:28] — The most common false assumption leaders make: “We need a perfect, complete AI solution before we can start”

[16:00] — Why you have to walk before you run: Building AI fluency before getting creative

[18:50] — Culture of curiosity as a prerequisite — and the operating model questions nobody wants to answer

[22:10] — The 5 organizational prerequisites for scalable AI: strategy alignment, cross-functional collaboration, end-user proficiency, scalability/adaptability, and governance

[27:17] — Real-world example: How misaligned incentives killed an AI sales tool before it could work

[29:22] — The “died on the vine” persona: Organizations with a track record of investments going nowhere

[35:02] — Small teams, big thinking: Why modular pods outperform hierarchies in AI implementation

[41:26] — How SaaS vendors can shift from selling features to enabling customer value creation

[45:05] — Budget misallocation: Chasing the “keeping up with the Joneses” technology trap

[48:10] — The 3-stage AI investment framework: Experiment → Production → Scale with clear business cases at each gate

[54:30] — Upskilling for AI: Hands-on training in the context of actual work beats corporate e-learning every time

[55:42] — The busyness trap: AI is making people work more, not less — and that needs to be examined

Tweetable Quotes

“The question isn't what's the next new AI tool. It's what are you trying to be as an organization?” — Justin Trombold

“Coating everything with AI doesn't get you to the key problems. It just gets you a lot of slop.” — Justin Trombold

“AI is everything and nothing at the same time. That's what makes it so different from every other SaaS tool.” — Justin Trombold

“You can't solve complicated equations until you learn the basics of arithmetic. AI is no different.” — Justin Trombold

“Start small but think big. Get the right group of people invested and empowered — then figure out what scaling looks like.” — Justin Trombold

“The shift SaaS vendors need to make: stop focusing on features and functionality, and start focusing on customer value creation.” — Justin Trombold

“Generative AI is a forcing mechanism to take a step back and look at what you actually do.” — Justin Trombold

“Upskilling for AI has to be hands-on, and ideally hands-on in the context of work people are already doing.” — Justin Trombold

SaaS Leadership Lessons

1. First Principles Before First Tools Don't start your AI strategy with a tool evaluation — start with a clear problem statement. Deconstruct what your organization is actually trying to accomplish, then work backward to determine whether and how AI fits. Leaders who skip this step end up with impressive-looking dashboards and underwhelming results.

2. Perfection Paralysis Will Kill Your AI Initiative The biggest false assumption leaders make is that they need a complete, enterprise-grade AI solution before they can move forward. Waiting for the perfect solution is the same as staying seated instead of learning to stand. Start where you are, build fluency, and iterate.

3. Your Operating Model Is the Real Bottleneck Technology is rarely the limiting factor. Cross-functional collaboration, decision-making structures, end-user proficiency, and governance frameworks are what determine whether AI creates value or collects dust. Address the operating model even though nobody wants to.

4. Align Incentives Before You Automate One of the most expensive mistakes: deploying an AI-powered sales tool when your comp structure rewards customer retention, not new logo acquisition. The tool can't fight the incentive. Before you automate a process, make sure the human systems around it are pointed in the same direction.

5. Move Deliberately from Experiment to Production to Scale Successful AI organizations don't just run pilots — they have clear decision gates. What metrics justify moving from experiment to production? What economics need to hold for scaling to make sense? Build this framework early. Scaling AI isn't free, and more volume doesn't automatically mean more value.

6. SaaS Vendors Must Become Value-Creation Partners The companies that win in the AI era won't just sell licenses — they'll help customers understand what needs to be true outside their product for the product to work. Customer stickiness is declining. The SaaS vendors who invest in their customers' readiness and outcomes will build durable competitive advantage.