Matt Gittleman | JHH VC
Venture Wishlist shares ideas and themes VCs want to fund. Brought to you by Purpose Built venture studio.
Matt leads venture investing at JHH vc, a Family Office that invests in 10+ early stage companies each year. Previously, he was an investor at Blu Venture Investors and Core Capital Partners. Matt has held marketing and operations roles at LivingSocial, Catalant Technologies, and Yesware, and holds an MBA from Georgetown.
Matt’s venture Wishlist:
ℹ️ Aggregating public information: Matt has invested in companies like GenLogs that uses sensor data and computer vision to track truck movements and Cloverleaf AI that aggregates town hall and public meeting data.
🚚 Supply chain optimization: Capturing digital exhaust and actively digitizing supply chain data creates opportunities for AI-powered digital transformation of the supply chain.
🌊 Data collection moats: These could come from the ability to collect evanescent data (e.g. public live streams that disappear otherwise), with collection supply constraints (e.g. fixed data collection points), or scale
🧑💼 AI products that focus on middle managers: these managers are “front line” buyers for productivity-enhancing AI products, and often have decision authority for shorter sales cycles than traditional enterprise products.
Other insights:
🚦 “Vertical” data collection can serve “horizontal” customer use cases: GenLogs and Cloverleaf both aggregate a single type of data that is valuable to different industries, with varying use cases.
🔭 Projecting implications of general purpose LLMs: Part of Matt’s focus on novel data collection is to ensure that GPT 5, 6, and 7 increase rather than erode the value of data-based companies.
🔗 Combination of data sources: Some companies can create value by combining novel data sources with existing public or commercially available data, such as in out of home advertising.
Full interview
How would you describe the focus of most of the investing you're doing today?
Unlike funds that develop a fixed thesis for three to five years, we can be more opportunistic and adapt our thesis as the market changes.
One area we're particularly interested in is hyper-contextualized data and its strategic use versus mere data acquisition. We're looking at companies that gather public domain information in innovative ways. This can involve using edge computing to capture sensor data or leveraging publicly available datasets combined with proprietary internal or customer data. The goal is to deliver and maximize value to end customers.
We're not just talking about building large language models for the future, though that's a possibility. Instead, we focus on how companies use existing data sources to create significant outcomes for their customers.
I'd love you to share a few examples that illustrate the types of data and opportunities you've seen.
One of my favorite companies in our portfolio is GenLogs. They're based in Virginia and are building a national network of sensor data along highways. They're targeting billboard operators and employing computer vision algorithms to track truck movements across the country.
What's fascinating about this is their innovative use of public domain information. Instead of placing people on every bridge to inefficiently monitor truck traffic, they've identified 100 to 300 prime locations nationwide to monitor truck movements effectively. This addresses a significant problem in the shipping and trucking industry, where 35% of trucks are driven empty. Improving efficiency here benefits greenhouse emissions, shippers, brokers, border control, hedge funds, and more by ensuring trucks are filled to capacity and operating efficiently.
On a completely different industry spectrum, we have Cloverleaf AI, based in Denver. They aggregate live stream links from town hall meetings across the country and use content analysis tools to understand the discussions. This information is valuable to various entities: if you care about net neutrality, it helps track relevant discussions; for a construction company, it provides leads on potential projects like pothole repairs; for Airbnb, it monitors HOA regulations. This data is valuable to everyone from local politicians to Fortune 500 companies.
We also have a company in the out-of-home advertising space, OneScreen AI, building the largest inventory of such advertising. They leverage publicly available information, sometimes for a fee, combined with proprietary internal data, to optimize return on ad spend. In an era of GDPR, the right to be forgotten, and CCPA, out-of-home advertising is becoming more attractive due to fewer online advertising restrictions.
I'm curious if you've thought about what makes for a good data source?
A key factor is how defensible and hard-to-acquire the data is. For example, with GenLogs and Cloverleaf AI, the defensibility comes from unique challenges in data collection.
For GenLogs, a freight logistics company, their moat is the strategic placement of cameras and sensors. There are few locations that make sense at scale for building this dataset, which creates a significant barrier to entry.
In the case of Cloverleaf AI, while the town hall meeting links are publicly available, aggregating them is labor-intensive. The value they've built comes from the significant man-hours invested in this process. In the future, scraping tools or generative AI capabilities might streamline this, but Cloverleaf AI's first-mover advantage and contextual knowledge give them a head start.
Additionally, understanding how to build and present data solutions to the right buyers early on is crucial. This involves developing alert systems and recommendation engines tailored to their needs. It's about contextualizing the information not just within its immediate environment but also in a broader macroeconomic or geopolitical landscape. Combining this data with other third-party sources can provide unique insights that competitors can't easily replicate.
Are there any other macro themes that you've been exploring?
What I like about this theme is its cross-industry applicability. Although I haven't done much healthcare investing, I see companies solving transparency issues around costs and improving patient outcomes by leveraging publicly available information combined with internal data from organizations like Red Cross or Blue Cross. This approach creates improved patient outcomes and is an area I'm interested in.
Looking at GenLogs, they use edge sensor data to monitor truck movements. I've seen companies applying similar technology for public safety and traffic management in public spaces.
Regarding other specific themes, it's impossible to ignore AI. The challenge investors face is determining the long-term value of AI companies as new versions of generative AI platforms, like GPT-6 or GPT-7, emerge. Will these companies improve or become obsolete?
I focus on companies creating more efficient processes and delivering value quickly with verticalized AI models tailored to specific industries. For example, we invested in a company optimizing the government grants process, which is traditionally labor-intensive. Some companies have reached over a million in ARR in nine months, leading to high valuations due to rapid growth.
I was curious if there are specific customer profiles you think are either non-consensus or underserved that you would point founders to explore more.
There is a lot of focus on selling to high-level executives or using a product-led growth strategy targeting engineers and end users who then advocate for the product within the organization.
I think some of the most critical problems are being solved by middle managers. These managers are often understaffed and looking for ways to enhance their team's capabilities and resources.
Particularly with AI resources, middle managers can significantly benefit by turning a team of three into the equivalent of a team of five or more through automation and other enhancements. This approach can be more effective than aiming for enterprise-wide agreements, which can be challenging to secure. Instead, focusing on these middle managers, who may have easier sign-off authority, can facilitate more agile, high velocity sales.
This strategy differs from traditional enterprise sales and product sales that rely on single-license purchases. For example, Cloverleaf has had some success in this area as a business development tool, targeting these middle management levels. While I can't speak for their long-term strategy, they have demonstrated the potential of this approach.
Are there any other big shifts or compelling catalysts for significant opportunities that stand out to you?
Supply chain optimization is a big area of opportunity right now. Many have been searching for the next frontier of digital transformation, moving from pen and paper to a digital future. The inefficiencies within national and global supply chains are monumental. Companies that find ways to integrate data and automation can create substantial efficiencies.
One key principle I learned from CRMs like Salesforce is "garbage in, garbage out." Creating better, more consistent, and automated inputs will drive efficiency, especially in the supply chain space, but likely in other areas as well.
This interview has been edited for length and clarity.