2. Snowflake Inc. (NYSE:SNOW)
Number of Hedge Fund Investors in Q1 2024: 73
Analyst Average Share Price Target: $206.66
Upside: 52%
Snowflake Inc. (NYSE:SNOW) is one of the biggest cloud computing companies in the world. It is a data processing cloud company that allows businesses to consolidate their data and generate insights. Estimates show that Snowflake Inc. (NYSE:SNOW) holds a 22% market share in the data warehousing market, making it unsurprising that the firm’s latest market capitalization sits at $44 billion. While this relieves some of the pressure on management to pursue the holy grail of growth in the cloud industry, Snowflake Inc. (NYSE:SNOW) is yet to turn a profit despite its heft. This also increases the pressure on management to achieve profitability, or high free cash flow and operating margins. Snowflake Inc. (NYSE:SNOW)’s management had also set lofty expectations for growth, but when it withdrew its FY2029 $10 billion revenue target in February, investors reacted and the shares plummeted by 27% in the aftermath. The following quarter wasn’t great on the cost front either, as its gross profit and operating margin guidance for the full year were guided at 75% and 3%, respectively. These were lower than the 75% and 6% that it had previously guided.
However, Snowflake Inc. (NYSE:SNOW)’s management is quite optimistic about its AI initiatives. During the Q1 2025 earnings call, it shared:
Like first and foremost, I think it is important for all of us to acknowledge that AI language models are going to have an impact at multiple levels of what you can think of as a data stack. So for example, the way in which people are going to be migrating from an old system, an on-prem system to something like Snowflake, is going to be aided by the presence of a Copilot that can do much of the translation. We already have such a translation product and we think AI is going to make that go even faster. But in other areas like data cleansing, data engineering that are perhaps not as sexy, but nevertheless required a huge amount of investment in order to make sure that the data is enterprise grade.
We think AI is going to play a big role both in the creation of those pipelines, but also in things like how does one make sure that the data is clean. For example, if PII accidentally flips into a table or a distribution goes very wonky, language models can help detect deviations from patterns. And then going up the stack, we have a very acclaimed product for writing SQL, our Copilot within our user interface, that can significantly accelerate in analysts’ ability to get to know a data set and be productive with it. And then, of course, to something like a data API, which now begins to put enterprise data into the hands of a business user, but with a very high degree of reliability. And so my point is there is a broad impact. And I think things like automating some of the work that an analyst has to do, for example, to troubleshoot problems, will be things that a language model can do.
Having said that, for a variety of problems, small models, which we are perfectly capable of developing from scratch like we have done for document AI or more a midsized model like what we did with Arctic, actually suffices for the vast majority of the applications that I’m talking about. And so there are academic benchmarks like there’s one called MMLU, it’s a notoriously difficult benchmark, and depends very much on model size and how many dollars people are throwing at training those models. We can get a huge amount done with a small team under modest investment without needing to play at that level where you’re talking — companies are talking about spending billions of dollars. I don’t think we need to be there. I think being very focused on what we need to deliver for our customers will take us a long way with the amount of investments that we are making.
And finally, I will add that we have amazing partnerships with a ton of people. Even today, I wrote about how we’re collaborating with landing that AI and doing company, but we have partnerships with Mistral, with Reika with a ton of other companies. The field of AI is so large that I don’t think there’s going to be one company that is going to make every model that every person is going to use. We are very good at developing the models that we need in our core and we actively collaborate with a large set of players for other kinds of models. And obviously, they see value in the 10,000 customers we have and being able to go to market together. And so I think this is likely to continue for the indefinite future in terms of what we need to do.
