AI is reshaping B2B sales: Procurement agents and 10% monthly automation targets

Apr 16, 2026 · The GTMnow Podcast
🎧 PodShort 31 min squeezed to 3 Ai sprinklerAI Sales Tech
Episode artwork
Sam
Founder and CEO at TestBox
The GTMnow Podcast
31 min squeezed to 3
Full episode from The GTMnow Podcast
Quotable Moments

The way that people buy software has changed dramatically, very quickly. I think we all know the stats that people are spending 70 to 80% of their decision has already been made before they even speak to you.

SamFounder and CEO at TestBox
06:10

At TestBox, the expectation is that everyone is automating 10% of their job every month. So that in about 10 months, less than 12 months, your entire job is fundamentally different and you're not doing the things you were doing before.

SamFounder and CEO at TestBox
15:20

Within the next three to five years, all procurement will be done almost exclusively by agents, including the negotiation.

SamFounder and CEO at TestBox
21:00
Key Insights
  • The way people buy software has changed dramatically and very quickly, with 90%+ of purchase decisions ending with companies on the day-one list, which typically only includes 3-4 vendors, now shrinking to 1-3 vendors.
  • The first sales call has shifted from discovery for both parties to primarily discovery for the seller, as buyers have already done extensive research and are further down the sales cycle.
  • Buyers are now expecting to test products earlier themselves to build trust and confidence, especially with AI tools allowing them to get 70-80% of the functionality of a dedicated solution, making the 'trust and confidence gap' for vendors much higher.
  • The AE (Account Executive) role is fundamentally different now, requiring them to be avid users of AI tools like Claude Code and Cursor, automating 10% of their job every month, and expanding into other functions like go-to-market engineering.
  • The world is clearly moving towards more agent-to-agent interaction in the buying process, where procurement agents will evaluate software based on cold-hearted rubrics and technical capabilities, rather than human-to-human interactions.
  • Within the next 3-5 years, almost all procurement will be done exclusively by agents, including negotiation, with humans primarily discerning trade-offs and ensuring support when issues arise.
  • Companies need to foster a culture of experimentation with AI, setting goals like running 15 AI-related experiments per week across all departments, and actively encouraging employees to build and share their AI-driven solutions.
  • Leaders must take ownership of their team's adoption of AI, providing tools and training, and having tough conversations with those unwilling to adapt, to avoid future layoffs due to a lack of necessary skills.
Metrics Mentioned
  • 90%+ (Percentage of purchase decisions that end with companies on the day-one list.)
  • 3-4 vendors (Typical number of vendors on a day-one list, now shrinking to 1-3.)
  • 70-80% (Percentage of functionality buyers can get from general AI tools compared to dedicated solutions.)
  • 10% (Expected percentage of job automation for every employee each month at TestBox.)
  • 1.2x to 1.7x (Increase in account executive quotas over the last 12-24 months.)
  • 15 experiments per week (Goal for AI-related experiments run across TestBox.)
  • $30,000 per month (Hypothetical token usage Sam would prefer for his team, compared to $20 per month, to encourage AI adoption.)
  • $100 (Cost for a billboard in the SF Bay Area (excluding photography) as an example of creative marketing.)

RevBots.ai View:

  • AI Sprinkler stage teams must mandate 10% monthly automation to avoid obsolescence.
  • ARM-stage orgs should prep for agent-to-agent procurement by structuring product data for AI evaluation.
  • Tab Hoppers risk extinction if they don't adopt AI tools for buyer research and outreach.
  • SaaS Hoarders must consolidate tools into AI-native platforms to enable automation targets.
🎧Full Episode:The GTMnow Podcast →