One month ago, Google announced a set of partnerships with seven commercial open source providers. Among those announced was DataStax, which held its annual conference this year and, for the first time, an analyst day. While DataStax and the open source project it is based on, Cassandra, are differentiated on a technical basis, the company also represents an interesting contrast with its peers directionally both among the newly minted Google partners and more broadly.
Of the seven commercial open source partners Google announced, for example, DataStax is one of two along with InfluxData that has not introduced a non-open source, hybrid license as a means of protecting itself from competition from the cloud providers. This is not, notably, because the company doesn’t seem them as a threat; asked about who the competition was in the analyst sessions, the CEO of DataStax candidly acknowledged that the company’s primary competitive focus was not on premise competition such as Oracle, but cloud-based managed services offerings.
A few years ago, this would not have been particularly notable. Amidst today’s heightened tensions, however, it represents something of a divergence from others in market.
To put this direction in perspective, it’s worth considering five themes RedMonk consistently discusses with its clients broadly speaking, but our commercial open source clients in particular.
- Developers Matter:
For a firm known for its advocacy of developers as The New Kingmakers, this will not be a particularly surprising assertion but it is the foundation on which everything else rests. As has been discussed at length elsewhere, and most recently last week with a room full of open source project maintainers at a client event in Boston, a variety of factors have collectively contributed to the reality that developers and other practitioners have immensely more influence than they did a decade or more ago. If not for advocacy and usage by developers, in fact, the market for open source would be a fraction of the commercial market today instead of its default assumption.
There’s a reason that AWS became the phenomenon it has, and that reason according to Andy Jassy is developers.
Convenience is a Killer Feature:
If developers matter, convenience is by default a killer feature because developers value it. No developer used EC2, for example, in its initial incarnations because it was more reliable or more powerful than a hardware based alternative; it was neither. Developers used EC2 instead because it was convenient. They could have an instance up in ninety seconds without having to even talk to IT, and pay for it as they went on an as needed basis instead of as an up front capital expense.
The same logic has applied to countless other widely adopted technologies; Linux was easier to obtain than Windows, MySQL was in the Linux repositories and thus easier to obtain than even PostgreSQL let alone Oracle, and so on.
What many open source projects have struggled to grasp, however, is that just as open source won in many settings not because it was better but because it was more convenient than proprietary alternatives, so too are cloud-based managed services more convenient in many cases than commercially supported open source offerings.
The obvious implication, therefore, is that commercial open source organizations need to compete not purely on an engineering basis, but on the axis of convenience as well. Which effectively means that they need managed service versions of their products.
The Competition is the Cloud:
For many years, certainly longer than was logical strategically, a majority of commercial open source providers viewed their primary competition as existing on premise vendors. Oracle was the target for open source databases, WebLogic and WebSphere for open source application platforms, and so on. This was a problematic approach for many reasons, not least of which was that it ran contrary to the mounting evidence that developers were influencing procurement.
Developers, after all, had minimal affinity for the enterprise-oriented commercial incumbents because that technology had been built not for them, but for the CIOs they worked for. As a result, most enterprise products devalued the developer experience at the expense of CIO priorities like integration, scalability, and so on.
It is only in recent years that commercial open source organizations have awoken to the reality – and, arguably, over-rotated on the idea – that their primary competition is in fact cloud providers. Cloud providers whose nature is to offer, as noted, a broader array of products in a more convenient package than has been typical for commercial open source vendors.
Model Problems Won’t Be Solved by Licenses:
One increasingly common response on the part of commercial open source providers to the perceived threat from cloud vendors has been to relicense their existing product. Whether the original license was permissive or reciprocal, a number of commercial entities who author open source software have sought to thread a needle by still maintaining an open source-like model but one which prohibits usage of the software by the cloud vendors the authors are competing with.
Because these types of carveouts are fundamentally incompatible with generally accepted principles of open source software, the resulting licenses have created a problematic gray area which is not open source but can be misinterpreted – or worse, mischaracterized – as such.
But while these licenses represent a clear and present danger for open source communities, the real problem for the companies that employ them is that they’re attempting to use a license to solve a problem with the business model.
The demand from a customer standpoint and therefore revenue growth is increasingly going to be found in managed service offerings, as evidenced by MongoDB’s Atlas product achieving a $100M annual run rate in less than three years. The appropriate strategic realization, therefore, is to accept what Chef has: that source code is not the asset to protect, and certainly not at the cost of customer, developer and partner goodwill. Experiences such as MongoDB’s which is facing direct competition from Microsoft’s CosmosDB and AWS’ DocumentDB in spite of a radical license change suggests, in fact, that source code is less important than the APIs they instantiate.
Commercial open source providers instead need to understand that licenses are an inappropriate and inadequate remedy for what is fundamentally a model problem, and focus on one that provides customers with the managed services they want while relying on their expertise writing the original code to outcompete would be profiteers. Database-as-a-Service (DBaaS) models can be difficult and expensive for on premise software companies to spin up, but among other benefits they significantly ease the inherent tensions of open source software-only approaches.
Data is a Model:
For vendors – both open source and otherwise – one of the critical realizations to running a managed services business is that an entirely new business model is enabled. Specifically, usage of the telemetry aggregated from hundreds of customers or tens of thousands of instances can be analyzed to yield insights that no single customer’s data can. It’s no coincindence that two of the open source companies beginning to leverage this aggregated telemetry stream – DataStax and Red Hat – call their respective products by that exact name: Insights. They are examples of a product type predicted in this space many years ago, one that has only recently become more palatable to conservative enterprise customers.
Even as traditional on premise software license models are challenged and give way to cloud based alternatives – another subject that has been discussed at length elsewhere – brand new revenue streams become possible. Progressive software and service vendors are thinking not just how to deliver software to customers, but how to identify patterns from all of the other customers running that software to give users something they’ve never had before: suggestions on how to leverage that software more efficiently, whether in the operational or business sense of that word.
If we assume, even is just for the sake of argument, that those are common characteristics of and planning assumptions for successful commercial open source organizations, the stategy outlined by DataStax at its event is logical and appropriate for the market context moving forward. Specifically:
- Developers Matter: DataStax’s CEO, at least, took pains to note in front of an audience not traditionally noted for its developer affinity – that being industry analysts – that developers were key to its strategy.
- Convenience is a Killer Feature: Key to winning developers in turn was making DataStax more “easy and obvious,” not to mention more accessible to developer populations, the primary manifestion of which was the Constellation product.
- The Competition is the Cloud: To better compete with the cloud it believes is its primary competition, DataStax has both partnered in the case of GCP and stood up a standalone product in Constellation to ensure that users in search of cloud-based managed services projects have that option.
- Model Problems Won’t Be Solved by Licenses: To date, at least, DataStax has not attempted to resolve its competitive issues with cloud offerings with a relicensed product. It remains reliant on an open core model, which has inevitably resulted in community friction but does not require the company to blur lines between open and non-open code.
- Data is a Model: DataStax Insights, as mentioned, is a product explicitly built upon a data-oriented business model.
The correct strategy, of course, is no guarantee of success. It’s also worth noting that DataStax is late to market with its DBaaS offering; many of its peers have had as-a-service solutions in market for several years, the most visible of which is perhaps MongoDB’s Atlas. The opportunity cost of that delay may yet make itself apparent.
While DataStax has much yet to prove with its ability to execute on its vision, however, their path forward is likely to be lower friction than competitors which are battling not only market competition but would be allies in the open source world at the same time.
Disclosure: DataStax paid for T&E to its event and is a RedMonk client, as are Amazon, Google, InfluxDB, Microsoft, MongoDB, Oracle and Red Hat. Chef is not currently a RedMonk client.