Nov 17, 2023 · 33m · saastr
Who Will Win the Go-To-Market AI Race? with Stage 2 Capital Co-founder Mark Roberge
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Stage 2 Capital co-founder Mark Roberge explores how generative artificial intelligence is transforming B2B software, drawing on historical tech disruptions to outline tactical go-to-market, product moat, and beachhead strategies for agile startups.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
Jason as informed peer 0.0 Guest teaching 0.0 Guest disagreement 1.3 Jason pushing back 0.0
speaking balance: gold is Jason, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 9:00 Incumbent seat-based model dismissal
Roberge directly rejects the viability of copilot add-ons by incumbents, asserting their seat-based business models fundamentally block them from embracing true displacement.
Hardest push from Jason ▶ 13:00 Venture capital obsession critiqueSpeaking as a venture capitalist, Roberge pushes back against standard founder dogma by emphasizing that more wealth is created in non-VC backed businesses.
Biggest teaching moment ▶ 19:40 Diminishing returns of training dataRoberge walks the audience through why elite AI engineering talent at startups can defeat incumbent data moats because of diminishing model performance gains.
Jason holds their own ▶ 25:35 Stage 2 enterprise pilot findingsRoberge presents proprietary data from Stage 2 Capital's experiment showing how enterprise legal friction creates a temporary tactical advantage for startup adopters.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Jason as informed peer | Guest teaching | Guest disagreement | Jason pushing back | Why |
|---|---|---|---|---|---|---|
| Reimagining B2B Software on the AI Blank Canvas | 0 | 0 | 1 | 0 | This is a solo keynote presentation with no host interaction. Mark Roberge opens by drawing parallels between the 1997 internet boom and modern AI, urging founders to rethink their startups on a blank canvas. | |
| Navigating the AI Hype Cycle and Dot-Com Parallels | 0 | 0 | 2 | 0 | Roberge warns that early AI wrappers risk irrelevance, comparing current market dynamics to 1998 dot-com iterations like Netscape and AltaVista. No host participation occurs. | |
| The Innovator's Dilemma and Disruption in Sales Tech | 0 | 0 | 2 | 0 | Roberge applies Clayton Christensen's Innovator's Dilemma to modern sales tech, arguing seat-based incumbents will be hesitant to deploy full SDR replacement tools. No host is present on stage. | |
| Economic Disruption and Ethical Responsibilities of AI | 0 | 0 | 1 | 0 | Roberge explores the macroeconomic and societal risks of AI displacement before detailing why startups must design big but start with narrow beachheads like Amazon and Apple did. No host interaction. | |
| AI Moats: Proprietary Data, Talent Bake-Offs, and Partnerships | 0 | 0 | 2 | 0 | Roberge analyzes AI defensibility, arguing top engineering talent at startups can outpace incumbents with vast data due to diminishing model returns. No host is active. | |
| Stage 2 Capital Experiment: Enterprise Hurdles and Startup Opportunities | 0 | 0 | 1 | 0 | Roberge recounts Stage 2 Capital's matchmaking experiment where enterprise legal and IT bottlenecks blocked AI pilots, creating an opening for startups targeting other tech startups first. No host input. | |
| Tactical Go-To-Market Playbooks from Stage 2 Capital | 0 | 0 | 0 | 0 | Roberge outlines concrete go-to-market resources and frameworks produced by Stage 2 Capital partners including annual planning and scaling science. Monologue format. | |
| Why History Favors Nimble Startup Disruption | 0 | 0 | 1 | 0 | Roberge concludes his address by inspiring founders with Fortune 500 turnover statistics, framing agile startups as nimble fighter jets against incumbent aircraft carriers. |
Statements from this episode (12)
Roberge: AI is the first blank canvas for B2B software in 20 years
“For the first time in 20 years, it's a blank canvas. It's complete artistic freedom. Especially in our area of B to B software. We can re-imagine everything.”
Roberge: Startups face irrelevance from AI like 1997 on-premise software
“I think there's a decent chance that whatever you're working on as a startup, as a use case, as a problem, It might be irrelevant. It might be like it's 1997, and you're working on a client server, on premise piece of software, and you're about to get the rug …”
Roberge: Current AI Landscape Parallels the Internet in 1998
“And if we were to say what year this is in the internet rollout, it's probably around 1998.”
Roberge: Embedding AI into existing workflows is not true disruption
“Certainly iterative technologies, just embedding an AI in the workflows we do today. That's not the answer. That's not the disruption.”
Roberge: Petitions to pause AI will let bad actors pull ahead
“This technology is coming no matter what. I appreciate petitions to sign to try to stop it, but I, I'm not sure that that's the solution. It just seems like bad actors will be able to get a head, a leg on us, and I'd rather this technology be run by the people…”
Roberge: Non-VC startups generate more wealth than VC-backed startups
“More wealth is generated in startups in non-VC backed startups than VC backed startups.”
Roberge: AI startups should target areas where incumbents lack proprietary data
“If you want to stack the deck in your favor, choose use cases as an entrepreneur where the big company incumbents don't have the data. That'll stack the deck in your favor.”
Roberge: Top AI talent beats mediocre teams with trillions of data records
“So if you had to choose between having an A plus AI tech talent team with millions of records of data, and you were going against a mediocre AI tech talent team with trillions of data that the startup wins. And that starts to make sense with the diminishing re…”
Roberge: Data acquisition will become a core BD function for AI startups
“The acquisition of data becomes an important function probably within the business development domain. Cause we don't have any data.”
Roberge: AI founders should target startups first despite high churn
“If I'm gonna build my MVP, I'm gonna build it with the startup tech community. That's, like, scary for me to say, because the startup tech by itself, as an, that, that has a high failure rate, so you're gonna have a lot of churn, but it's just where you're gon…”
Roberge: Start annual planning by October 1 to hit revenue targets
“Don't wait till January to build your bottoms up plan. Cause you'll realize that you should have hired four reps in Q four to hit your 20, 24 plan. Please start your annual planning process on October first so you can hit your 20, 24 number.”