In the age of COVID, however, it hasn’t been enough for analysts to rely on traditional sources, as most of which aren’t updated very frequently.
Instead, it has been crucial to find alternative datasets that do a better job explaining what’s happening with the economy in real time.
And while that’s not exactly something new – a corporate jet data feed tipped off analysts to the Anadarko-Occidental Petroleum deal last year – it has become so important that entire cottage industries have sprouted up around it.
As a result, alternative data is starting to become commoditized. On Amazon Web Services’ marketplace alone, there are 982 large datasets available for free.
Spoiler alert: you’ve already seen some of them, as I am a relatively frequent user.
Now, the problem I have encountered most often is that they tend to be enormous, messy, and unstructured. And while some of that untangling has already been done by others, lots of coverage gaps remain.
The size of some of these datasets also makes it prohibitive – sometimes impossible – to process or analyze quickly in simpler spreadsheet programs like Excel.
A few wouldn’t even fit on my hard drive.
Because of that, there is a clear need for scalable, integrated data storage and external computing power that is only going to get bigger over time.
And to explain why, it’s easiest for me to use an example from my own professional career.
Rows, Rows, Rows Your Boat
One problem I’ve encountered in all my previous jobs is that analysts who gather or evaluate data don’t tend to simultaneously think about things like structure, storage architecture, or how to interconnect the information.
That’s not a dig on analysts, it’s just the nature of the job focus.
Whether I was making visits to iron ore mines in Brazil, copper mines in Chile, coal mines in Alabama or steel plants in Pennsylvania, my attention was entirely dedicated to learning about the operation and gathering primary data.
Once I got back to the desk, I would organize it, input it into our cost models…and send it off to others.
Over time, those “others” would grow to include both colleagues who worked within the same company, and advisors or consultants from external firms.
So, in the instances where those databases were huge and the external and internal colleagues organized them differently, it was almost impossible for me to interface them…which is the entire point of analysis.
Even just trying to pull together smaller datasets to work with on my desktop was a challenge, because filtering through billions and billions of rows of data to find something useful is incredibly slow.
As it just so happens, these are exactly the issues that Snowflake’s services aim to solve.
And although research firm Gartner now projects global IT spending to fall by 8% in 2020, the work-from-home nature of the pandemic is causing a 6.3% increase in spending on cloud networks.
More importantly, within that cloud segment that SaaS is by far the largest single component.
I’ve written at some length before about the struggles of the commercial real estate market during this pandemic. As work-from-home has become more accepted, the need for companies to hold long-term leases – or even require offices in general – may be permanently on the decline.
Because of that, it stands to reason that with a decrease in lease square footage also comes a decrease in on-premises tech infrastructure for data storage and processing.
But in no way does it remove the demand for products that can provide those same services… that demand will shift to the cloud, causing an already-fast growing segment of the market to accelerate even more.
And what Snowflake has essentially done is set up the market’s only one-stop shop for all of those needs.
More “Brisk and Useful” Than “Beautiful and Unique”
In short, Snowflake provides “warehousing as a service” that consists of three primary components – database storage, query processing, and cloud services.
The storage itself is backed by Amazon S3 – their scalable cloud storage platform – under Snowflake’s account, where it is encrypted, compressed, distributed, and optimized for performance.
The query processing takes place in “virtual warehouses” which are essentially clusters of computing power available to extract data from storage, and to process or filter it.
And finally, cloud services handle all other tasks such as authentication, security, and query compilation.
But the really attractive part is that all of these features can be scaled up or scaled down as needed, and customers only pay for what they use.
Moreover, it’s not just the only focused, pure-play data warehousing company on the market, but it’s also the most versatile.
While all their competitors – Amazon’s Redshift, Google’s Big Query, and Microsoft’s Synapse – are enormous, they all lock in the user to using their native cloud service.
Snowflake, on the other hand, is the only software in existence that can run on all three.
And as data-driven businesses small and large alike – or even individual analysts like myself – shift toward work-from-home setups that require flexibility of both cost and function, I see Snowflake as likely to be the biggest beneficiary of the fastest-growing segment of the market.
Yes, the +120% runaway IPO valuation yesterday was insane.
But the rampant criticisms and references to tulipmania I have seen from the financial pundits in the Twitterverse have mostly come from folks who clearly aren’t handling their own large datastreams.
I think this company will crush it.
And while 238 times its projected revenue (~US$264 million) for this year is not exactly a price I want to pay right now, I do want to own it. After all, if they 10x revenue in the next three years – which is completely possible in our new COVID world – that multiple falls to 23.8, which is frankly conservative for the tech sector these days.
So much like a certain TV pundit said on-air the other day, wait for the next tech pullback – which we all know is coming – and start picking up some, little, by little… until you’ve made a little snowball.
Just watch out for avalanches!
All the best,