Unlike other infrastructure categories, data is a relatively new department at most companies—most of today’s data teams, let alone the “Modern Data Stack” they’re using, didn’t exist during the Global Financial Crisis. This gives us little historical precedent on which to base our understanding of how data will fare through a recession.
Founders and data execs alike have long argued that a thoughtful investment in data will help companies do more with less. As IT budgets contract, we predict that argument will be put to the test, and as a result, create the blueprint for a more streamlined, more effective data stack.
Driven in part by VC excitement, the data stack has exploded over the past 10 years. As companies began centralizing on cloud data warehouses, a slew of tools emerged to aggregate, manage, and make use of data. Some of these new companies reinvented old categories (e.g. Snowflake in data warehousing, Looker in BI, Fivetran in ETL) with better/faster/cheaper value props, while others invented new categories (e.g. Monte Carlo* in data observability, dbt in transformation/metrics, Hightouch in reverse ETL) with net new ROI propositions.
The deluge of new options for data leaders has dramatically improved how data teams operate, but it has also created noise in the market. If you ask two data leaders what their stacks look like, you’ll likely end up with very different answers. Vendors, more than buyers, have owned the narrative around how you should construct your data stack, creating confusion around where to start. As IT budgets tighten, we expect that to shift.
We’re beginning to see data leaders approach existing tools with more scrutiny on pricing, and new tools with more scrutiny and time to value and ROI. Open source companies are facing higher hurdles to justify the cost of their cloud alternatives, and vendors are having to make business cases where they previously focused on technical advantages.
A tougher selling environment, while painful, will ultimately accelerate the pace at which the “Modern Data Stack” matures. As data leaders are forced to focus resources, they’ll need to separate the “must-haves” from the “nice-to-haves.” Pressure to solve pressing problems quickly and a shorter leash for tools with long implementation cycles should benefit the best teams.
An added focus on driving business value with fewer resources should force conclusive best practices to emerge more quickly than they otherwise would have. This will create some big winners in the “must-haves” category, many of which already have mindshare today. It should also pave the way for the next set of data founders, who will have a more established stack on which to build and with which to integrate. We believe the net effect will be that data teams emerge more critical to their companies, and despite the pain along the way, data companies see the same benefit.
*Represents a company in GGV’s portfolio
Read more VC predictions: