It was a busy four days at the Snowflake Summit 2022 last week. Now, I'm the first to admit that I'm not the most data-tech-savvy person on our team. But I am good at understanding the business side of things, so I thought to collect my insights from the Summit for other non-techies interested in the topic. Enjoy!
First and foremost, I was pleasantly surprised by how many keynotes and sessions were easy to follow for a non-techie. Mind you; I happily skipped all the labs and deed dives! But more than that, the core of many of the sessions was about real-life (business) problems and how we can solve them. Often with technology, but primarily by understanding the industry domain and task at hand. And by changing the thinking, behaviors, and processes of the people involved.
Snowflake's focus on industry solutions is one clear example of the shift from technological solutions to business problems. Frank Slootman, the Snowflake CEO, talked in his keynote about how lately, more and more customers start conversations: "I'm not interested in technical details, but I do have this problem. How can you help me fix it?" It's not enough that you are excellent at software business: you also need to understand your customer's business deeply. I've witnessed this shift on the software consulting side. It was interesting to hear that the same is true for "pure" technology companies.
Snowflake's answer to providing end-to-end solutions to its customers is a robust ecosystem with skilled partners and best-of-breed solutions working well together. Solita and we at Agile Data Engine have first-hand experience with what it is like to be a solution and technology partner (Elite and Select, respectively) [1]. Talking with some of the Powered by Snowflake partners at the Summit, it was clear that Snowflake is investing much time and effort to help them out.
Technology and consulting companies — or the sharp tools and minds that use them — have always cooperated to win the customers' hearts. With the era of monoliths long gone, the trick is to choose a set of tools that work well together. Some technology companies have opted for a closed system, where they provide all tools by themselves. Snowflake has gone the other route and welcomes compliant solutions to be part of its ecosystem. This open ecosystem provides opportunities for smaller players to join the big leagues.
Where I think Snowflake is looking to change — not just the game but — the playing field is the Marketplace. The ability to purchase access to data products from third parties, or Data Sharing in Snowflake's world [2], provides the customers a unique opportunity to enrich their data relatively simply. Once you get your customers consuming data only available on your platform, it becomes much harder for them to leave your ecosystem.
At the Summit, Snowflake launched the Native Application Framework [3], enabling developers to build and monetize applications on the Data Cloud. Watching the demo that combined Snowpark for Python [4], Streamlit [5], and native apps, I could see how smooth the developer experience for simple ML-related apps might be. Getting a pre-trained ML model with a basic interface to crunch my numbers for optimal marketing campaign spend does sound appealing!
For me, the democratization of data means that regular folks should have two things: an understanding of how data works and access to relevant data. Data democratization is vital since data is already fueling the next wave of innovation. Diversity, equality, and inclusion are also important aspects to consider.
I'm a massive fan of not reinventing the wheel. So providing access to cleaned and ready-to-use high-quality data products make a lot of sense. And I think if you make an app that somebody else would benefit from, it is not just smart but also sustainable to share that app with the world.
There is a catch, however, at least in my mind. How can you be sure you can trust the data product that you buy? Or the app that somebody somewhere made? Trust is an underlying issue of any platform economy. As a platform owner, you can either provide curated and validated products from carefully selected partners or trust that peer review will weed out the bad seeds.
With the first approach, your ecosystem is only partially open, and you spend great effort validating the products and solutions. Once your platform grows big enough, the sheer volume of products makes it hard for a specific partner to stand out in the crowd. Lack of visibility will wreak havoc on the motivation of these selected partners who have made considerable investments to be part of your ecosystem. In addition, one might also ask whether the platform delivers on the promise of democratization if there is a for-profit organization curating what gets served and to whom.
The second approach seems to be the more democratic one. Let the users of your marketplace test out the products and decide if they are any good and deserve a five-star rating. But when I, the customer, build my business on these products, I kind of don't want to be the one testing them out, right? With data and digital solutions at the heart of my business, I need to have 100 % trust in my foundation. Some app that some teenager somewhere hastily copied from somebody else to make a quick buck is not something to build a business on. And I certainly don't want to enrich my data with a fabricated dataset created with malicious intent.
I, for one, am eager to see how Snowflake's ecosystem evolves. I love that they are not settling for the role of just a cloud data warehouse provider and want to "help mobilize the world's data." [6] I can see how all these steps and announcements fit into that vision. Now they only have to win our hearts and convince our minds to trust them and the Marketplace.
Lastly, to answer the question mentioned in the title of this blog post, Snowflake talks about powering the data economy in a recent trending article. [7] So yeah, I guess they are building a platform economy for data products.
[1] Solita becomes Snowflake Elite service partner
[2] Snowflake for Data Sharing
[3] Snowflake Native Applications
[4] Snowpark Python: Bringing Enterprise-Grade Python Innovation to the Data Cloud
[5] Democratizing Data Apps — Snowflake to Acquire Streamlit
[6] Snowflake: About the Company
[7] Snowflake: Supporting the Data Economy