Metaplane’s fund raise details. It’s been great to join the company as an investor and team member during the round! I’m looking forward to seeing how high we can fly over the coming months.
As excited as I am about the announcement of the round, I want to step back a bit and take a dispassionate look at the segment (as much as this is possible).
When Monte Carlo raised in May of last year, people were saying things like “Data Observability is recession-proof”. I jumped on that bandwagon and I still believe that this should be true (yes I used the s-word, sorry for my potty mouth). I still believe that because data quality is a fundamental to data success, come rain or shine, Data Observability is a necessary part of achieving data quality on any data stack that has production data.
However, it was possibly incorrect to link the timing of the announcement of Monte Carlo’s raise to it being recession-proof, given when it probably took place. Monte Carlo probably raised long before the May announcement - my experience with the delay between Metaplane raising the actual cash, and announcing it, is that there can be a gap of many months. The market really went bad at the start of 2022 and it’s possible Monte Carlo actually raised before this.
Does this mean that Data Observability isn’t, in fact, “recession-proof” and attractive to investors in a bad market? In short, no. Metaplane is an example of an org where investors are still willing to invest well after the market shifted, and possibly invested when most people in the VC world thought the sky was falling. And it would seem that it’s not an exception - while I was trying to raise money in Q2 2022 for Avora, a VC mistook Avora for a Data Observability company (it’s a Metrics/Business Observability company) and were very keen to move fast, until I corrected them.
Should VCs still be looking to invest in Data Observability startups? In short, no. Part of why I joined Metaplane is that I looked at Metaplane’s approach to GTM, and thought it would be very hard for new and existing competitors to compete. If you think about how any company would need to compete, they would have to compete in one or more of the following ways and possibly more than one at the same time:
Price - with multiple OSS offerings in the market and PLG SaaS like Metaplane, it would be hard to come into the market and compete aggressively on price.
Integrations - the later-funded offerings in the market have a huge number of integrations and the engineers to maintain them. It’s very hard to compete with a larger tech company in this way.
Features - fundamentally, how current generation Data Observability tools work is very similar, and this is why Metaplane can compete with the series stage companies in the market for some of the same deals. There are nuances and bolt-ons to the core feature, which is anomaly detection on a metadata time series trend. But make no mistake - this is the core feature.
Enhanced lineage acts as a force multiplier for this core feature, which is why you see many of the existing products offering column level lineage and integrations with ELT/Orchestrators/Reverse-ELT/BI tools. This could be a way for a new entrant to succeed - to provide a truly best-in-class and complete understanding of data lineage, where alerts are made and gathered in this context.
I think if you were entering the market now, you’d be a bit late to compete with the existing companies in offering the same features. It’d really be like taking a knife to a gunfight. On top of that, we will also see some consolidation and pivots in the space over the coming months.
This isn’t to say that a company that has been in stealth mode, building a great offering for a year or two, couldn’t come in and win business… they could, but starting to build a product from scratch NOW to perform a similar role to the others out there is unlikely to succeed.
Where is west?
There are more spaces to invest in Data Quality tooling - I’ve mentioned some of these in my series on Data Contracts. However, given the size of the problem, I think there is room for more ideas and entrants. They also have a large market to disrupt: for the CDPs (I use this term loosely, but I mean companies like Snowplow, Rudderstack, Amplitude, Segment et al) the big problem plaguing their data is quality - garbage in garbage out has never applied more readily to any product segment.
A new generation of Data Contract/schema validation/just an API companies could emerge, which focuses on ensuring that data being collected is valid and complete (see my post below for full details on what I mean by this), that simply sends the data to a data warehouse or onwards to a system that needs real-time events. They could reflect that the gravity of data has moved towards the warehouse and therefore be simpler in architecture and therefore cheaper. They could also focus more on enhancing the Software/Data/Analytics Engineer developer experience, to protect data quality during development. Data sovereignty has been a big problem too, many of the CDPs have stored and processed data, asking customers to pay to access their own data!
If I were to invest in Data Quality tooling, as a VC, today, this is where I’d want to put my money.
Come join my subscriber chat, where there is thread to discuss our raise and market.