Revenue advisor and ex-CRO says collaboration with AI beats pure automation in sales processes
Dave Govan explains why sales teams get better results by collaborating with AI throughout the sales process while keeping human judgment in every key decision.
Human Judgment: Human-in-the-loop AI accelerates sales work while leaders retain responsibility for discovery, relationships, pricing, and final decisions.
Pipeline Focus: AI-driven targeting and account research help sales teams prioritize stronger opportunities and build qualified pipelines faster.
Workflow Context: Shared AI workspaces preserve account history, meeting transcripts, and decisions, improving analysis throughout complex sales cycles.
Reality Check: AI delivers leverage rather than autonomy; unattended outputs remain generic or inaccurate and require clear human oversight.
CRO Leadership: Successful AI transformations rebuild revenue workflows around collaboration, measurable execution, and human judgment instead of simply buying licenses.
Dave Govan has a history of driving revenue at organizations like Oracle, VeriSign, Hitachi, and Dell RSA. He is the founder of G2 Strategic Advisory Services, where he advises technology CEOs, investors, and revenue teams on go-to-market strategy.
We sat down with Dave to learn how human-in-the-loop collaboration with AI benefits sales processes more than pure automation. Here's what he said.
Turnarounds and breakouts
Dave's Thoughts
My path to this AI moment is not new…The companies that win the next decade will not be the ones with access to AI. Everyone has access. They’ll be the ones whose leaders and teams change how they execute. That gap is exactly where I operate.
I'm the Founder and Senior Principal of G2 Strategic Advisory Services. I advise technology CEOs, investors, and revenue teams on go-to-market strategy and on implementing AI within the revenue engine. I do this as an operator, not a consultant. Across my career, I've driven $2.3 billion in revenue as a CRO and sales leader, served on six executive leadership teams, and scaled companies at every stage from seed to public to PE-owned.
Create a free account to continue reading this article.
Want more from the CRO Club? Sign up to get free access and join modern revenue leaders using proven playbooks, peer insights, and AI-driven strategies to build predictable pipeline and scalable growth.
I advise revenue organizations, but my foundation is enterprise field sales leadership. Early in my career, I led major accounts regions as a field RVP at Oracle, generating over $100 million. That grounding in large, complex enterprise selling has shaped how I've built every organization since.
At the enterprise and PE-owned stage, I led Dell RSA's Americas sales organization, a 250-person team of generalists, specialists, channels, and inside sales, growing revenue from $400 million to $540 million. I was then promoted to Chief Revenue Officer of NetWitness, the cybersecurity business that was carved out and sold to private equity. There, I led a 500-person global organization and lifted EBITDA 52 percent. We sold enterprise security software with services to security operations teams at large enterprises and government agencies.
At the startup stage, I built and led the revenue organizations at Sailthru and Dynamic Yield, AI and big-data martech companies. We sold SaaS personalization to digital marketers in e-commerce and media. There, the work was zero-to-scale: building the team, playbook, and GTM motion from a few reps into a worldwide organization, selling to CMOs and their teams.
My path to this AI moment is not new. A decade ago, I was already working in AI and big-data martech, scaling personalization platforms like Sailthru and Dynamic Yield. I've watched AI move from a feature you sold to the operating layer to human-in-the-loop collaboration.
The companies that win the next decade will not be the ones with access to AI. Everyone has access. They'll be the ones whose leaders and teams change how they execute. That gap is exactly where I operate.
How AI reshapes account targeting and pipeline management
I made AI the operating layer of my own go-to-market, starting with how I target and pursue accounts.
I now run an AI-driven targeting and research motion. I use AI account-intelligence and intent signals to identify which accounts are in-market and worth my time, and I collaborate with AI across the pipeline to research buyers, build go-to-market maps, and draft client materials. I stay the human in the loop on judgment, positioning, and every client-facing decision.
As a result, leverage and signal improved. I run a larger and better-qualified pipeline than I could by hand, I reach a credible point of view on a prospect in a fraction of the time, and I produce work at a depth that used to take a team. I also help my clients make the same shift, which keeps the advice honest. I am not selling a transformation I have not run on myself.
How AI is used in the sales process
I'll walk you through the sales process from a fresh lead to a signed order, and how AI supports it.
I recommend creating a Claude Project for each new opportunity and each signed client. A Claude Project functions like a workspace where you can collaborate with Claude. You can share relevant files and prompts. It will remember all of your interactions, which is great for collaboration. Your job is to teach it and leverage it, but you own the opportunity and outcome. Trust your mind more than the responses and keep redirecting it when you need to.
…organize your discovery call with AI. Shape the agenda and the questions to make the conversation sharp and earn the right to go deeper. The two questions that anchor it never change: “What do you need done?” and “What does success look like to you?” The discovery call itself is human.
Start with pre-call research. Once a lead is obtained, you can use AI applications like Clearitty.com to obtain account intelligence on the company, the buyer, the likely pressures, and the gap between their stated needs and probable needs. You can expand your research to also include frontier models like Claude. Starting with an informed perspective and point of view is better than know-nothing, generic discovery.
Then, organize your discovery call with AI. Shape the agenda and the questions to make the conversation sharp and earn the right to go deeper. The two questions that anchor it never change: "What do you need done?" and "What does success look like to you?"
The discovery call itself is human. But an AI notetaker like Granola.ai can provide a transcript that you can give to your AI project. In doing so, you capture the full context of what the buyer said and can use that information, rather than letting it sit in your memory or on a notepad. I suggest doing the same for every meeting with the account and feeding that back into the Claude Project. As the sales cycle goes on, you can ask it to examine all of the transcripts and provide insights or guidance.
Dave's Thoughts
…you are the decision maker and need to trust your gut instinct as well. Your mind is better suited for creating winning strategies. AI will often miss context unless you tell it.
When you enter the proposal and negotiation stages, you can collaborate with AI to shape positioning and sequencing, leading with what they asked for and then opening the larger opportunity once trust is established. Test different scenarios. Ask it to assess how competitive you will be for this type of project, etc. Pressure-test the commercial structure and pricing to ensure they hold up under scrutiny, and use AI to identify where you may be overcomplicating the deal or leaving money on the table. However, you are the decision maker and need to trust your gut instinct as well. Your mind is better suited for creating winning strategies. AI will often miss context unless you tell it.
Then, AI builds the quote, using your own templates, in a fraction of the time it previously took by hand. You bring the judgment on scope and terms.
All these steps occur within an AI workspace, rather than through different one-off prompts. This means the workspace retains context and can make better inferences as additional questions or concerns arise.
You run the sale, the discovery, the positioning, the close. AI makes you faster and sharper at every step, captures everything so nothing gets lost, and helps you along the way.
How specificity decreases AI drift
Occasional drift occurs in my AI workflows, but we resolve it with simple redirects. The more you use it, the less it happens.
To avoid drift, always be precise about what you are doing and why. Write explicit prompts providing the full picture of the sales scenario. The more complete the prompt is, the less back and forth you’ll have.
Why revenue processes need a human in the loop
Overall, I lean on AI to power analytical and production work where speed and breadth win: account targeting and prioritization, buyer and account research, market mapping, ICP and buying-committee work, and synthesizing extensive input into a clear point of view. AI should be pushed hardest into pipeline prioritization, deal scoring, forecasting, and churn-risk signals because the data supports it.
But while AI informs and accelerates decisions, it never makes the decision.
Humans explicitly handle judgment under ambiguity and anything built on trust.
Discovery stays human. Understanding what a buyer needs and what success looks like to them is a conversation, not a query.
The relationship and the close stay human, including quieter calls, such as deciding what to surface or withhold from a buyer, and when to pursue or pause.
Pricing and commercial terms stay human. AI often recommends pricing that I know will not work because, at scale, it fails the buy-versus-build or buy-versus-hire test. When reminded of that, AI then gives a different answer.
Even with AI-assisted forecasting, I tell teams to maintain human override for the final number.
Overall, I see it as a collaborative human-in-the-loop process, and humans own quality and outcome.
Why revenue leaders need to correct their unrealistic expectations
I have not had a case where AI failed to deliver the revenue impact I expected. Since AI is software, I never expected it to deliver revenue. I expected leverage, and it has delivered that.
The one place reality fell short of the hype, however, is autonomy. The idea is that you point AI at a workflow, and it runs on its own. In practice, unattended output is generic, sometimes confidently wrong, and still needs people in the loop. It makes us faster, not absent.
I would not call that AI underdelivering. I would call it people needing to correct unrealistic expectations.
They need to understand that human-in-the-loop collaboration with AI is the way to go. Why? AI is amazing software that excels at supporting thinking, but not at thinking itself.
Why MCPs are crucial for software businesses
The Model Context Protocol (MCP) enables any software company to connect with any frontier model/LLM through connectors, combining the best of both worlds. That means your competitors are likely allowing users to interact with their applications via Claude, ChatGPT, Copilot, and Gemini.
If your offering does not embrace MCP, you risk going out of business.
How to improve sales productivity
Here are three changes you can make to significantly improve sales productivity, pipeline conversion efficiency, and forecasting accuracy:
Use Claude more collaboratively.
Use AI signals to prioritize target accounts and target contacts. I'm working with AI technology from Clearitty.com that excels at this. I've redesigned top-of-the-funnel execution.
In larger organizations, I suggest using Clari.com to monitor opportunity and account health for more accurate forecasting.
How CROs can lead successful AI-driven transformations
And here are three pieces of advice for CROs integrating AI into their workflows: First, the gap is not access; it is collaboration. Second, lead this from the operator’s seat — not as a tech rollout. Third, keep judgment human. The CROs who win this moment will not be the ones with the most AI. They will be the ones whose teams changed how they execute and how effectively they collaborate with AI.
And here are three pieces of advice for CROs integrating AI into their workflows:
First, the gap is not access; it is collaboration. Everyone says they use AI, but very few effectively collaborate with it. Your teams likely already have the tools. You must rebuild the workflow around these tools and learn to work with AI effectively for anything to change. Buying licenses is not a strategy; changing how work gets done is.
Second, lead this from the operator's seat — not as a tech rollout. Your job is not to make people excited about AI; it is to show them where it makes real work faster, sharper, and better, and where it does not. Stay grounded in execution and revenue impact, not abstract AI capabilities. Do not pressure people with "fall behind" framing; it backfires.
Third, keep judgment human. AI excels at compressing analytical work and widening your perspective, but decisions where being wrong is expensive, or where the value lies in the relationship itself, still belong to a person.
The CROs who win this moment will not be the ones with the most AI. They will be the ones whose teams changed how they execute and how effectively they collaborate with AI.
