Jan 25, 2016 · 28m · mad
Investing in Data and A.I. // Dan Scholnick, Trinity Ventures (Hosted by FirstMark Capital)
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
In this Data Driven NYC session hosted by FirstMark's Matt Turck, Trinity Ventures General Partner Dan Scholnick discusses his venture capital strategy for data and AI, the rise of developer-focused sales models, and key growth lessons from portfolio companies like New Relic.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Matt holds 15.4% of the talking time here. How this is scored →
Matt as informed peer 2.8 Guest teaching 3.8 Guest disagreement 2.1 Matt pushing back 1.1
speaking balance: gold is Matt, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 0:05 Dan dismisses New York market priority to a room of NY founders
Dan bluntly tells the New York tech community that NY is not a primary focus for Trinity and notes that they moved portfolio company InfluxDB to San Francisco.
Hardest push from Matt ▶ 0:10 Matt challenges Dan on automation timeframeWhen Dan asserts that eventually everything can be automated, Matt pushes back directly by challenging whether that applies to the next 5 years rather than 50.
Biggest teaching moment ▶ 0:21 Dan dismantles the power law revenue assumption in developer toolsDan corrects an audience member's assumption about power law distribution by explaining that developer tools achieve a far more democratized revenue model spread across SMBs.
Matt holds his own ▶ 0:12 Matt synthesizes AI landscape models to challenge guest on investment strategyMatt demonstrates high domain familiarity by synthesizing a previous speaker's talk and setting up the horizontal versus vertical AI debate for Dan.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Matt as informed peer | Guest teaching | Guest disagreement | Matt pushing back | Why |
|---|---|---|---|---|---|---|
| Data Driven NYC Event Graphic | 3 | 2 | 2 | 1 | Matt introduces Dan and highlights several of his portfolio investments, while Dan lightheartedly jokes about VC motives over dinner. Dan clarifies that Docker is a foundational technology rather than a pure data company. | |
| Trinity Ventures Strategy and New York Market Perspective | 2 | 4 | 4 | 2 | Dan candidly tells the New York tech audience that New York is not a focus area for Trinity and mentions relocating portfolio company InfluxDB to San Francisco. Matt humorously notes that Dan is losing the room and prompts him to explain his reasoning. | |
| Investment Themes: Practical Problem Solving and AI | 2 | 3 | 1 | 0 | Dan outlines his investment thesis around problem-solving data tools and the ongoing inflection point in deep learning. Matt asks open-ended questions to allow Dan to elaborate on his portfolio choices. | |
| The Necessity of Human-in-the-Loop AI Systems | 4 | 4 | 2 | 2 | Matt pushes Dan to clarify the timeframe for AI automation versus human-in-the-loop systems. Dan explains why training data needs and 80 percent model accuracy limits require human involvement for the foreseeable future. | |
| Horizontal versus Vertical AI Investment Focus | 5 | 4 | 2 | 2 | Matt frames questions citing earlier event speaker concepts on horizontal versus vertical AI, as well as classic VC skepticism around developer tools. Dan explains how power has shifted from IT groups to developers, enabling superior direct-to-developer business models. | |
| Overhyped Categories and Undifferentiated Data Tools | 3 | 3 | 3 | 1 | Matt prompts Dan to share cold takes on overhyped categories. Dan dismisses traditional enterprise software and undifferentiated horizontal data analytics tools. | |
| Audience Q&A: Revenue Distribution and Market Democratization | 0 | 5 | 2 | 0 | During audience Q&A, Dan reframes a question about power law revenue distribution, explaining that direct-to-developer SaaS businesses actually democratize revenue across thousands of SMBs rather than relying on an 80/20 top-heavy concentration. | |
| Key Lessons from New Relic's Growth Trajectory | 3 | 5 | 1 | 1 | Matt asks Dan to share key takeaways from New Relic's rapid growth trajectory. Dan delivers detailed lessons on forcing product quality by delaying sales hires and looking for non-obvious domain experience when scaling. |
Statements from this episode (16)
Scholnick: Docker is foundational tech for big data developers
“Yeah, it's less of a pure data company, but it's certainly a foundational technology that's making it easier for developers to work with big data systems.”
Scholnick: Trinity Ventures closed a $400M 12th fund
“We just closed our 12th fund, which is a four hundred million dollar fund.”
Scholnick: Trinity GPs serve on no more than eight or nine boards
“We have more partners per dollars in our fund than any firm in Silicon Valley, and that means that no, no general partner at Trinity's on more than eight or nine boards, and that, so that means we do the work for our companies, and we don't rely on other peopl…”
Scholnick: Trinity invested in InfluxDB and moved them to San Francisco
“I invested in InfluxDB, and I think it's a great company, but we moved them to we moved them to San Francisco.”
Scholnick: New York is not a focus area for Trinity Ventures
“To be, I mean, to be truly honest with you New York isn't a focus area for us.”
Scholnick: Most tech industry value is still created in the Bay Area
“Still most of the value being created in the tech industry is in the Bay Area.”
Scholnick: AI commercialization will replicate the massive enterprise boom of Big Data
“And to me it feels like, Big data. Maybe six or seven years ago where companies were real waking up and realizing we have all these data assets. We need to do something with them. And that led to the rise of Hadoop and the Hadoop vendors and then, you know, a …”
Scholnick: Humans will remain heavily involved in ML for 5-10 years
“So, at least for you know, I think the next five to 10 years, there's going to be heavy involvement of humans in machine learning systems.”
Scholnick: Vertical AI will be more successful near-term than horizontal platforms
“But that's, I think unless we see more crowd flowers, the vertical approach is going to be a lot more successful in the near term because it won't rely on the customer to have expertise in how to do things like tune models and tune training sets and things lik…”
Scholnick: Tech organization influence has shifted from IT groups to developers
“The power in, and the influence in technology organizations is, has moved, shifted dramatically away from the classic IT group to the developer.”
Scholnick: Trinity prefers developer-focused software to traditional enterprise models
“So, so anyway, we love these businesses at Trinity. I'm looking for more and more of these businesses and businesses, and I like them a lot better than the traditional enterprise software business.”
Scholnick: Hard to build a business on undifferentiated horizontal data tools
“There are so many undifferentiated products and services out there. And I think the advice that was given by the earlier speakers of solving problems, specific problems for the customer, instead of just trying to be Horizontal in, you know, we have a better qu…”
Scholnick: Data scientists dislike enterprise sales pitches as much as developers
“But I think you can build a similar business to The type of, to the direct to developer tool businesses. There's no reason why not. I don't think that data scientists enjoy being sold to any more than developers do.”
Scholnick: Developer tool companies rely less on large enterprise customers
“I've been surprised at New Relic and looking at financial results of Atlassian and other companies that I've seen in the developer space in that, you know, the old, I think this is what you mean by power law, but I, you know, the assumption was always that 80%…”