.png)
Podcast
March 20, 2026
Episode 85: This Should JUST Take 5 Minutes | A RevOps Rant Session
Two RevOps veterans rant about the most relatable chaos in operations: broken enrichment tools, scope creep, and the "this should take 5 minutes" myth.
What does it actually mean to keep up with AI as a RevOps professional; not the hype version, but the practical, roll-up-your-sleeves version? That question sits at the center of a refreshingly grounded conversation between co-host Camela Thompson and Taimoor Tariq, Founder of GTM Base, who brings 12 years of revenue operations experience spanning sales ops, Salesforce architecture, marketing ops, and CS ops to the table.
Taimoor's path into RevOps will sound familiar to many: he started in sales, complained loudly enough about broken processes that his manager told him to go fix them, and never looked back. His agency, GTM Base, now focuses on four core areas: CRM and tech stack, data foundations, AI and automation within go-to-market motions, and operational structure. What emerges from this conversation is a practical framework for how RevOps professionals should be thinking about AI right now: not as a magic wand, but as a powerful extension of the problem-solving muscle that drew most of them to the discipline in the first place.
Before diving into tools and tactics, Taimoor makes a point worth dwelling on. When someone once asked him why he stayed in RevOps, his answer surprised them.
"I find it so creative. I love solving problems and RevOps throughout my career gave me this opportunity that there's a new problem every quarter, every month, and we have to put our brains to solve it." — Taimoor Tariq
That framing matters because it shapes how he thinks about AI's role in the function. If RevOps is fundamentally about creative problem solving, then AI isn't a threat to the job. It's a force multiplier for the skill that defines it. The question isn't whether operators will be replaced, but whether they'll learn to deploy these tools as effectively as the problems they're solving demand.
Camela captures the mood in many RevOps teams right now with characteristic directness: it's not really fear of missing out, it's "fear of being fired." Executives are pressing teams to do more with less, and AI has become the answer to that pressure in the minds of leadership whether or not the foundations are actually in place to support it. This tension between executive expectation and operational reality is one that anyone trying to align RevOps strategy with tactical execution will recognize immediately.
One of the most important things Taimoor establishes early is that AI doesn't bypass the foundational work. It amplifies whatever you already have. Good data and well-structured processes become more powerful with AI; bad data and broken processes just fail faster.
"In order to have a good AI rollout map, you need to have good data, good processes. And that's what RevOps thrives on." — Taimoor Tariq
This point is worth sitting with. The classic RevOps failure mode of reaching for new technology before getting the fundamentals right applies equally to AI. Camela makes an astute observation: for years, the reflexive executive answer to operational problems was "we need a new CRM." Now it's "plug in AI." The underlying problem, skipping the hard diagnostic work, is the same. As Taimoor puts it, you need the data in the right structure in your CRM before AI can make it useful.
This isn't a reason to avoid AI. It's a reason to treat AI adoption as an opportunity to finally pressure-test the data foundations that have been accumulating technical debt for years. RevOps teams that do that work now will be significantly better positioned than those that bolt AI onto a shaky infrastructure. If you've been avoiding the data management conversation, this is the moment to have it.
When Camela asks Taimoor what RevOps professionals should be prioritizing right now, and what the one thing is if you're slammed, his answer has two distinct parts.
Taimoor's recommendation for practitioners ready to get their hands dirty is Claude Code, the agentic coding environment from Anthropic. His enthusiasm is specific and grounded in real usage.
"I feel like it's such a big, powerful technology that I wish I had access to throughout my entire career of RevOps." — Taimoor Tariq
He describes Claude Code as distinct from simply using ChatGPT or Claude in a chat interface. The key differences: sophisticated problem-solving capability (what he describes as "PhD-level problem solving"), the ability to actually execute tasks by accessing tools and data rather than just generating text, and the ability to write and run code. The result is something closer to having a team of engineers on call than having a smarter search engine.
His advice for getting started is deliberately anti-course: don't binge tutorials, just start using it on real work. "Get your hands dirty. Put it into your regular workflows. Once you start doing that, you'll get a hang of it more."
The second priority is less flashy but arguably more durable: getting genuinely good at RevOps fundamentals. Taimoor argues that AI has made execution cheaper and faster, which means the differentiator has shifted.
"What has shrunk now is the middle. The time it would take for you to actually implement a change has probably shrunk. That also means what you need to get better at how you actually set yourself apart. That means getting better at the beginning and at the end [outside of that implementation period]." — Taimoor Tariq
The "beginning" is requirements gathering, clarity of thought, and translating a sea of noise into the signal that actually matters. The "end" is change management, rollout, and the judgment calls that determine whether a technically correct solution actually gets adopted. These are the parts of a RevOps project that AI can't do for you, and they're increasingly where the real value gets created or destroyed. The broader conversation about moving from ticket taker to strategic operator has never been more relevant.
Camela raises a concern many practitioners share: what if experimenting with AI means accidentally writing bad data back into the CRM? Taimoor offers a practical framework for safe experimentation.
The core principle is separating the read environment from the write environment. Start with CSVs. Extract data from the CRM or data warehouse and store it locally, then run analysis locally without touching production systems. Operate in a sandbox, get confident in the output, and only then figure out how to push changes back into production.
This might seem obvious, but as Camela notes, the people who have lived through the trauma of a 4:00 AM call from a CEO know exactly why it's worth stating explicitly. Taimoor confirms he's had that call. Camela had a similar experience on her honeymoon. The lesson isn't "don't experiment"; it's "experiment with appropriate containment."
The parallel to good change management practice is direct: test thoroughly before deploying, and make sure you can roll back if something goes wrong.
To make these ideas concrete, Taimoor walks through a recent client engagement that illustrates what agentic AI actually looks like in practice on a RevOps problem.
The client operates in the e-commerce space and segments by revenue tier: enterprise, mid-market, and so on. The problem is that standard data providers have reasonable coverage at the top end but miss a significant portion of the long tail. Good companies were falling through the cracks because no third-party source had their revenue data.
The traditional approach would have involved weeks of research, vendor evaluation, and manual analysis; likely a month-long project requiring additional headcount or outside help. Instead, Taimoor's team used Claude Code to tackle it in an afternoon.
The process unfolded in phases. First, they gave Claude Code the problem and asked it to generate a research plan. It spawned multiple agents, did extensive research, and returned with the three best approaches for finding the missing data. Then the team extracted known customer data from the CRM — companies where revenue figures were available — and enriched it with market signals: Amazon ratings, review counts, marketplace presence, and e-commerce-specific data providers. Finally, Claude Code ran a regression analysis to identify which signals correlated most strongly with revenue, creating a proxy model that could fill the gaps.
"While we are doing this, somebody in my team asked, 'How did you use to do this before Claude Code?' And I was like..." — Taimoor Tariq
Camela's answer: "You didn't." That's the point. This isn't AI doing something faster that humans were already doing. It's unlocking analysis that practically wasn't feasible before. The implications for smarter segmentation and ICP refinement are significant.
A second use case Taimoor highlights is one that will resonate with anyone who spends time pulling data from multiple systems for regular reporting.
RevOps teams routinely stitch together metrics from Salesforce, HubSpot, data warehouses, support tools, and other sources. This kind of multi-system reporting is exactly the type of repetitive, structured work that burns cycles without building value. Taimoor's team has started using Claude Code to build custom Chrome extensions that automate this work directly in the browser.
The concept: the extension logs into the relevant tools, executes the manual tasks that would otherwise be done by hand, runs on a schedule, pulls the data into a structured format, and can even run analysis on the compiled output. What was a weekly reporting chore becomes a background process.
This connects to a broader principle Taimoor articulates: instead of asking "what are the cool AI tools?", ask "what are the manual workflows that are eating my team's time?" The use case should drive the tool choice, not the other way around. This is a mindset shift that matters, and it connects directly to the case for automation in document and workflow management that RevOps teams have been building for years.
The episode's title poses a binary, but Taimoor's actual answer is more nuanced and more useful than a simple yes or no.
His framing: a business has two core jobs. 1. Building products or services, and 2. selling them. Everything else supports those functions. RevOps didn't emerge from nothing; it emerged in large part because of CRM technology creating problems that needed to be solved at scale. Then Clay and the GTM engineering wave created a new layer of operational work that RevOps absorbed and evolved to meet.
The pattern suggests something important: the problems will always exist. How we solve them changes. What threatens practitioners isn't AI itself. It's getting married to specific tools and technologies rather than staying attached to the underlying problems.
"Let's not get married to tools and technology, but rather the problems." — Taimoor Tariq
The role won't disappear. It will evolve, as it always has. For operators worried about the pace of change, Taimoor offers something rare in AI conversations right now: genuine reassurance grounded in historical perspective rather than hype.
"This is a very exciting time. It might also be a really scary time for a lot. And I would say if you are someone who actually feels that things are moving so fast, you're not alone. I sometimes wake up in the middle of the night and feel that as well." — Taimoor Tariq
Check out our blog, join our community and subscribe to our YouTube Channel for more insights.
Our average member has more than 5 years of RevOps experience. That means you’ll have real-time access to seasoned professionals. All we ask is that you’re generous with your knowledge in return.
.jpg)
