Tools Of The Future: How AI Changes The Way We Interact With Data
What This Means for Investment Tech & Software More Broadly
👋 Hi, I’m Andre and welcome to my weekly newsletter, Data-driven VC. Every Tuesday, I publish “Insights” to digest the most relevant startup research & reports, and every Thursday, I publish “Essays” that cover hands-on insights about data-driven innovation & AI in VC. Follow along to understand how startup investing becomes more data-driven, why it matters, and what it means for you.
Current subscribers: 19,060, +435 since last week
Brought to you by Affinity - Insights you need to find, manage, and close more deals
In a survey of 700+ dealmakers, almost 90% predict they’ll do the same or more deals this year. But is this optimism warranted? Get the full picture in Affinity’s new report, including strategies that leading VCs are using to source more deals using their networks and how they’re implementing AI to streamline internal processes.
Why Our Tool Landscape Will Change
Last week, I wrote about Investment Tech as an emerging category. It’s software that facilitates public and private market investing for those firms that cannot afford - or simply don’t want - to develop their software stack in-house.
I received lots of feedback and questions on this post supporting my thesis: Most investors are lost with make vs buy and overwhelmed with the ever-growing spectrum of tools available - that btw all sound the same.
How can you find the right tools for your budget? Why will the 4th wave of tools be different? Is it smart to wait until it unfolds or jump on the train just now?
To answer these questions, I decided to dive into the evolving 4th wave of Investment Tech and highlight how AI-powered tools will be different from anything we know so far. In the end, I’ll also share what I would do as an investor looking to become as data-driven as possible with as few resources as necessary.
4 Waves of Investment Tech
Quick summary of the four major waves of Investment Tech:
1st wave: Private company and fund data providers like Crunchbase, Dealroom, Pitchbook, Preqin, and others
2nd wave: Systems of record (CRM, portfolio, cap table management) like Affinity, Vestberry, Carta, and others
3rd wave: Signal providers like Harmonic, Synaptic, and others
4th wave: AI Copilots and unified systems like … ?
While the first three waves had a great impact on “what” investors do, they didn’t materially change the “how”. I know that some of you would disagree, but let’s face it: Workflows and the ways how we interact with data are still mostly the same. Manual and inefficient.
Yes, we use commercial databases instead of Google. Yes, we use CRM systems instead of Excel files. Yes, we use signal providers instead of analyzing all data points and extracting the insights ourselves. But all in all, investment processes haven’t materially changed for the past decades. Until now.
Standing on the shoulders of giants (systems of record, data & signal providers), the 4th wave of Investment Tech will finally change the way how investors interact with data. This time it’s less about the what but more about the how.
Join 19,000+ thought leaders from VCs like a16z, Accel, Index, Sequoia, and more.
Status Quo: A Fragmented Tool Landscape
Today’s tool landscapes are fragmented. Period. This is as true for professional investors as it is for founders, operators, and any other knowledge workers.
Some interesting stats:
Overall spend per company on SaaS tools grew 50% from 2018 to 2020
Unique number of SaaS tools in usage per company grows by about 30% YoY
Number of different SaaS tools in usage across businesses:
Small businesses (102)
Mid-market businesses (137)
Isn’t this crazy? We went from a single generic operating system (think Microsoft and MS Office Suite) into a highly fragmented stack with tens and hundreds of expert tools for almost every task. The 1st, 2nd, and 3rd waves of Investment Tech contributed significantly to this fragmentation.
To show you that this is not just mere theory, I wrote about my investor tool stack here. I use 80-100 tools regularly. Eithy to one hundred 🤯 Where is this going?
Naturally, we sooner or later face a tradeoff: Increasing efficiency by adding another expert tool vs decreasing efficiency due to increased context switching across too many tools.
Two options to solve this: Either we just stop adding new tools and stick with our continuously aging stack or we continue to add new and kick out older tools. According to the study linked above, the latter causes companies to churn through 30% of their tools every year. This can’t be the future.
There’s Light: Learnings from ChatGPT
Why did ChatGPT pick up growth and adoption like nothing before? Because it’s powerful and easy to use. Suddenly, we had all of humanities knowledge at our fingertips. The magic lies in “making data actionable”.
Looking deeper into the ChatGPT case, there’s more we can learn and apply to the future of software more generally:
ChatGPT (Nov 2022) started centralized and generic, just like all software stacks with Microsoft OS and MS Office Suite just few decades ago —> THE PAST = centralized generic
Customized GPTs (Nov 2023) increased capabilities and expertise, yet led to a highly fragmented landscape, just like with software stacks (see my example of 80-100 tools above) —> TODAY = fragmented expert
GPT Store (Jan 2024) added one conversational layer on top of Customized GPTs, essentially an LLM interface that classifies your intent and roots it to the right Customized GPT in the back. It’s centralized but able to execute tasks on an expert level. Best of both worlds —> THE FUTURE = centralized expert
I described this in more detail in my ChatGPT review & outlook last year.
Translating the accelerated ChatGPT evolution to Investment Tech, I expect 3 main drivers for the 4th wave:
Deep integrations: Tools need to be connected via APIs. Few Investment Tech solutions today smoothly integrate. This needs to change. While most solutions are human-first and provide a web app and nice UI/UX, rich APIs and tool-to-tool connections will be key.
Single source of truth: Fragmented siloed data needs to be merged and stored in a single place. We need version control for data just as we do for software.
Making data actionable: Following deep connections across tools and datasets, we need a simple but powerful interface to interact with our data in natural language but at the same time dive deeper for tasks like KPI benchmarking, commercial due diligence, and more. AI and chat interfaces will be crucial in transforming data into insights.
While it took OpenAI just 14 months to move from (1) centralized generic ChatGPT over (2) fragmented expert Cutomized GPTs to (3) centralized expert GPT Store, it took software stacks more than two decades to move from (1) centralized generic OS to (2) fragmented expert tools (=best of breed). This is where most of us are stuck today.
The evolution has been slow and merging our highly fragmented tool landscape back together into (3) a smooth, unified expert stack might take some time again. The reason why OpenAI was able to compress this timeline into a fraction of the general software landscape is the fact that they have everything standardized under one umbrella.
This is different from today’s software landscape which is far from standardized. Moving ahead, we need more standardization across tools. How do tools communicate? How do they collect, ingest, process, and version data?
Today’s software stack is built for human-to-machine (=web apps and nice UI/UX) and machine-to-machine (APIs) communication. It’s not built for human-to-AI, AI-to-machine, or AI-to-AI communication.
To get there, we need to reinvent software communication. How rich are the endpoints? Where are the limits for AI? Do we need human-in-the-loop checkpoints? These and many more questions need to be answered before we move ahead to a state like the GPT Store.
“Software eats the world but AI will eat software.” To get there, however, we need to reinvent the software stack and answer the questions above. For Investment Tech, I expect that some of the 1st, 2nd, and 3rd wave players will push in this direction with the ambition to become the single source of truth with unified versioned data that can be accessed through an AI Copilot interface.
I believe that most incumbents will struggle to drive this change as the 4th wave requires fundamental disruptions that go beyond adding new features or an AI interface. It requires rethinking the fundamentals of software stacks to eventually unify everything smoothly under one umbrella. A great opportunity for new entrants.
To close the loop on the initial questions of “How can you find the right tools for your budget? Is it smart to wait until it unfolds or jump on the train just now?”, I’d currently personally rather wait on the sidelines and keep a close eye on emerging players. I expect that the software landscape for investors but also more generally for knowledge workers will dramatically change in the next year or so.
Today, it’s difficult to predict whether incumbents can drive this change or whether it will be the new entrants capturing this opportunity. In any case, you don’t want to be stuck with legacy stacks for too long. Your competitors aren’t sleeping.
To keep you in the loop and help you cut through the noise, we’ll highlight some exciting “4th wave of Investment Tech providers” in our upcoming “Data-driven VC Landscape 2024”. Take our short survey here to shape the story and be the first to receive the full report, including all insights about modern investor tool stacks.
Thank you for reading. If you liked it, share it with your friends, colleagues, and everyone interested in data-driven innovation. Subscribe below and follow me on LinkedIn or Twitter to never miss data-driven VC updates again.
If you have any suggestions, want me to feature an article, research, your tech stack or list a job, hit me up! I would love to include it in my next edition😎