Kamesh Pemmaraju is the Founder and Managing Partner of OptimaGTM, a GTM systems consulting firm, focused on a specific and urgent problem: B2B and AI startups unable to reach predictable growth/scale after seeing some initial market traction.
In our conversation, Kamesh shares his diagnosis of why early traction stops working, why the instinct to add headcount or tools rarely solves the real problem, and what it actually takes to achieve repeatable growth in an investing environment that has shifted decisively toward GTM efficiency over growth at all costs.
M.R. Rangaswami: What are you seeing break down for B2B and AI founders that are scaling to achieve repeatable growth, and what actually causes it?
Kamesh Pemmaraju: The pattern is almost identical across companies. They’ve hit $1M or $2M on founder-led sales, strong early relationships, genuine problem-market fit (initial validation of opportunity and demand). The instinct that got them there was right. Then the same instinct becomes the ceiling.
What breaks when it comes to scaling the business is that their Go-to-Market was never designed as a cohesive end-to-end system that connects their product decisions with sales, marketing, and customer success execution.
When functional leaders optimize locally without a shared system connecting them, decisions start to conflict instead of compound. Marketing generates activity that sales can’t convert. Product ships features that customers don’t adopt. The founder, who used to be the thread holding everything together, becomes the bottleneck instead.
By the time we’re brought in, the typical reaction has already been tried — another analytics tool, a new hire, a strategy offsite. None of it works because those are function-level responses to a system-level problem. You can’t fix a structural gap one department at a time.
M.R.: The B2B buying environment has changed significantly. How does that change what founders actually need to do differently?
Kamesh: Three things have shifted simultaneously, and most founders are only seeing one of them.
The first is structural. Buyers now spend only 17% of their time meeting with vendors. The rest happens independently — through peer forums, analyst reports, and AI-powered search — before a founder ever gets in the room. 43% of buyers make defensive purchase decisions 70% of the time. They’re not asking “is this the best product?” They’re asking “can I defend this choice if it goes wrong?” SQL-to-close rates have dropped 5-6 percentage points year over year as a result. The issue isn’t lead generation. It’s that founders are running persuasion playbooks against buyers who need validation.
The second is specific to AI products, and it’s the one most founders aren’t prepared for. Buyers are using AI themselves and are starting to deploy agents extensively. That changes the conversation in the room completely. They’ve moved past being impressed by “AI-powered.” They know enough about how these systems work to be genuinely skeptical of vendor claims. They want to know precisely where AI sits in the workflow and what it actually does. And the objection that didn’t exist three years ago is now one of the first ones you hear: “why can’t we just build this ourselves?”
The third is that novelty is gone. We’re in what ICONIQ calls the Execution Era. Buyers have moved from curiosity to ROI scrutiny. The demo doesn’t close the deal. Customer adoption data does. Case studies from companies in their exact vertical do. Being cited in an AI search when their procurement team researches the category does — because that signals market legitimacy without a vendor saying so.
When all three converge, the old playbook doesn’t just underperform. It actively works against you. The GTM system has to be redesigned around validation, not persuasion. That’s a structural change, not a messaging tweak.
M.R.: What does predictable revenue actually look like at this stage, and what separates the companies that get there from the ones that don’t?
Kamesh: Predictable revenue is not a forecast that happens to be accurate. It’s an output of a system that was designed to produce a specific result.
The shift starts with one decision: stop treating GTM as a series of campaigns and start treating it as a product with a lifecycle. Every great product goes through the same stages — defined, built, launched, measured, iterated, and expanded. GTM works exactly the same way, but most companies skip the first three steps entirely. They launch motion before they’ve defined who they serve, why they win, and how their product creates measurable value. Then they wonder why conversion doesn’t improve no matter how hard the team works.
Following the lifecycle means sequencing correctly. Define your ICP and positioning before building any motion. Design the full system — from demand generation through sales through customer expansion — before running it. Launch with enablement in place, not after. Measure at the system level: pipeline quality, win rates, customer time-to-value, CAC payback. Run a quarterly review to diagnose drift before it becomes expensive. And design explicitly for expansion — NRR and upsell don’t happen by accident, they have to be built into the system from the start.
The companies that get this right understand one thing: acquisition feeds retention, retention feeds expansion, and expansion feeds the next acquisition cycle. When those stages are connected and measured together, revenue becomes forecastable. When they’re treated as separate campaigns owned by separate functions, you get motion without momentum.
The founders who make it to $20M are not the ones who worked harder. They’re the ones who built the system first — and then ran it.
M.R. Rangaswami is the Co-Founder of Sandhill.com