The Starting Line

A post by Cack Wilhelm

There is a Silicon Valley maxim often shared with entrepreneurs entering their public debut: the IPO is not the conclusion. It is the start of life as a publicly-traded company.

My promotion to General Partner at IVP feels similar. It's every bit the culmination of my career to date, and the start of a new chapter.

I began my career in venture after working in enterprise sales roles at the data infrastructure companies Oracle and Cloudera. While my professional interests have expanded, my early affinity for investing in this sector has only deepened. As I look ahead, here are a few ideas I'm considering in the field of data infrastructure:

Best of Breed

AWS will take a page from Amazon (Retail): At a $71B+ run rate, AWS can't be the best at everything, a fact that even the company itself might concede. For instance, while AWS has both a Snowflake and a Databricks equivalent, it hasn't prevented either Snowflake or Databricks from growing into many-billions in enterprise value. AWS has already moved from what was once a competitive stance (where re:Invent announcements were feared), to one of co-opetition. I predict that AWS will continue in this direction as it moves toward a position that is outwardly cooperative. A perfect example of this is Amazon.com which started as a marketplace that sold first-party goods almost exclusively and today generates more than 50% of retail sales from third-party sellers.

It pays to fit-in: Outside of the big cloud platforms, each software vendor is increasingly required to integrate with ecosystem partners or adjacencies in order to reduce any friction to adopt. The value of intra-platform integration will decline as the possibility and prize for being best-of-breed increases. In my experience moving from a sales role at a company with a 13-page product sheet (Oracle) to one with a single SKU (Cloudera), I learned an appreciation for the broader ecosystem and the requirement to fit-in.

Web3 Infrastructure

Web3 infrastructure, or infra3, remains in its beginnings. The ecosystem still needs abstractions and modularity, specific and discrete building blocks atop the blockchain protocols, as well as the ability to connect each distinct block together.

Meeting with founders and researching companies in web3 and infra3 brings me back to my first days in venture circa 2013. Then, I knew a passable amount about data and data infrastructure from prior jobs. I remember leaving meetings with more questions than answers and obscure acronyms scribbled in the margins of my notepad for review.

I'm back there again today, only now, it's the margins of my Notion doc that are filled. This time, I'm better prepared: equipped with the knowledge from investing in data infrastructure over the past eight years. Crypto, after all, is just another type of database: a chain of blocks, a historical log, albeit of a different sort - one with a consensus protocol and a shared incentive structure.

As I find myself moving between the familiarity of web2 and the unfamiliar world of web3, I've come up with a few hypotheses:

Standardization will happen quickly. While there is plenty of Twitter debate regarding the future and efficacy of web3, it matters not for my purposes: if cryptocurrencies, NFTs, and web3 remain self-referential and exclusively focused on crypto use-cases, each application still requires underlying technology.

If Amazon Web Services offers more than 200 first-party services, the web3 ecosystem might have 20. Compute, storage, and wallets/identity are mostly resolved, but I predict near-term standardization around primitives for messaging, multi-chain wallets, oracles, monitoring, analytics, and security.

Building blocks will be built as byproducts. Just as Slack arose as a byproduct of the online game Glitch, I expect we'll see similar instances in the build out of web3. Web3 developers will set out to create games and dapps, only to realize that many of the dependencies have yet to be built. (Good examples of this are Dapper Labs or Immutable.)

Analytical capabilities: It still feels like we're in the early innings of web3's analytical capabilities. Blockchains store data and handle state but they are dissimilar from databases in not indexing the data, making it difficult to use as a result. We can query the blockchain, and we can reference others' work for analytics, but building up one's own analytical capabilities is still hard, especially for more complex queries. As more and more business cases become clear, solutions will likely follow.

Modern Data Stack

The Snowflake Effect: The demonstrable performance and enterprise value of Snowflake has reinforced data and cloud infrastructure's place on the map and resized the price (up). The extreme wealth generated for investors and employees alike re-cut the math, to where a 0.05% employee grant or a 4% investor stake seem worthwhile. Snowflake solves the question of where to store and process the data; I anticipate that we will see further innovation in moving data in, out, and between these data stores, and in software to observe and confirm that the data pipelines are accurate.

Persona sprawl: We will see a broadening of analytical infrastructure offerings, focused increasingly on data-driven business teams and software developers, on sub-second response times, and on production answers for end-users.

New(er) databases overcome inertia: Database evolution seems to take decades not years, but for a handful of emerging databases, those years are behind us. I anticipate that 2022 will be a break-through year for long-developed databases, whether they are time-series, serverless, NewSQL, or analytical. Having sold Oracle databases for a living, I remain partial to the Oracle database, but eve

LINK:	https://www.ivp.com/news/blog/the-starting-line/...
	See more stories from ivp

More from IVP