Snowflake's IPO - changing cloud computing... forever

Analytics as a true competitive advantage remains more myth than reality. Don’t get me wrong. I am totally convinced that some companies have redefined their business and turned their market upside down by leveraging world-class data and analytics platforms (e.g. Uber, Amazon, Facebook).

One of my favorite customers is a concrete manufacturer that pulls Internet-of-Things data off their trucks to ensure quality and on-time delivery. Missing a delivery can cost 10’s of thousands of dollars in lost revenue. For them, analytics truly drives the bottom line, in real-time.

So, the promise is there. The tech exists. However, traditional data warehousing principles, largely dictated by Moore’s Law, have been upside down and counterproductive for quite some time. While there are many players who have seen this coming, Snowflake is about to do to the data warehousing market what the iPhone did to Mobile.

Traditional data warehouse design has always been factored for performance. Only bring the data you need. Spread performance across tiers and stages. Create indexes and views to optimize what you think users want. And that performance has always been limited by storage and processing, which have been very expensive to consume with legacy hardware manufacturers.

The business impact of these constraints has been, simply put, an inability to ask and answer the critical second and third level questions around performance. Sure, we can easily see if sales are up or down, but how efficiently can we drill-down on those questions in efforts to enrich the real picture of what’s happening. Can we pull in the meta-data quickly to paint a more detailed picture? Can we connect financial data to sales data to supply chain to consumer confidence…?

The simple answer is ‘yes’, but in general that analysis is not happening real-time or in one system/database. More likely, that analysis will require multiple logins, multiple people, and often times a request to IT to create a new report or even a new data mart. Ninety days later we get our analysis, but by then its yesterday’s news.

Perhaps even more importantly, how much flexibility do your best business analysts have in doing data discovery. Do they have the tools and the data set necessary to follow the breadcrumbs? For instance, a customer of ours in the casino and hotel business deployed an analytics solution to look at all of their food and beverage pricing and profitability. While the solution provided great insight, the ROI came through a simple, daydreamers question. “How much do we charge for a cup of coffee across all of our kiosks?” It turned out that the price difference was big and correcting it single-handedly paid for the project. Does that type of analysis and insight sound possible to you?

To cut to the punchline: Cloud computing is much, much more than just putting your data in someone else’s basement. It is more akin to a ‘GoFundMe’ that is supported by 1,000’s of investors (customers) who all want to share the most powerful computing resource on earth. And furthermore, if you were to build such a “Great Brain” that you knew you could sign on 1’000’s of customers to pay for, you would design it differently. You would leverage cheap and redundant storage and build a massively parallel engine that breaks down mathematical problems and divides the work efficiently across a network of computing nodes. Well, that’s exactly what Amazon, Microsoft, and Google (among others) have done with their cloud platforms. And in turn, developers have started to build the tools to leverage the newly founded resources.

Here’s where Snowflake comes in. What they’ve done, so elegantly, is built a platform that brings the amazing power of cloud computing to data engineers and knowledge workers that doesn’t require them to relearn everything. Sure, the best consultants might be able to build a more bespoke and fancy solution for your company than Snowflake. Maybe... (but probably not cheaper).

But remember, the iPhone was not unique in that it invented all the technology required to make it work. Quite to the contrary. It took 100’s of innovations and packed them in an elegant, user-friendly package. And changed mobile forever by doing so.