Yesterday’s Slack and AWS partnership announcement feels very significant and raises some interesting questions. In short, Amazon employees will adopt Slack, Slack will migrate to AWS’s Chime calling infra for Slack Calls, and the companies will tie Slack and AWS closer together with integrations like AWS Key Management <> Slack EKM. The coverage of the announcement positions this move as a united front against Microsoft and its Teams onslaught, and while that is a valid interpretation, the rabbit hole goes a bit deeper.
In search of the enterprise workflow “iPhone” moment
Far from just being a story about the narrow competition in a single market, this AWS-Slack announcement is unfolding in the context of a very complex, overlapping set of battles and interlocking “friendpetitors.” Because of this, I am much less interested in the “Teams vs. Slack” or implications of this move and more about how it fits into the broader landscape. After a decade or two of the “consumerization of IT meme,” the workplace software space is heating up on a few distinct but related fronts.
- A fight for synchronous remote communication via voice and video calling is playing out in the midst of a pandemic, accelerating the remote work and communication trends already well underwayWith many white collar employees now working remotely far more (sometimes exclusively) than before, the strategic importance of the tool of choice for video communications has never been clearer. One only need look at Zoom’s stock price and user growth to understand why this is now a crucial space. There has also been an acceleration of thinking around what it means for video calls to be more than an app and more like a “platform.” Companies like RemoteHQ and Gong are experimenting with integrations and experiences that assume the existence of a video call as a backbone and build on top – like digital whiteboarding, agenda and minutes, or sales-intelligence. Zoom is trying to be much more than just a video chat app, and we can see the growth of Zoom’s annual developer conference and many of the apps built on Zoom’s APIs as a preview of this new “ecosystem” phase of the video chat market’s evolution
- A battle for the “workplace operating system,” centered around different theories for what the chokepoint/keystone asset in a work ecosystem will turn out to be.We have multiple different companies trying to build the defining company which will hold a differentiated and defensible position at the center of the emerging SaaS+Remote+Integrations work stack. Everyone is vying to be “The Platform” that everyone needs to plug into, everyone needs to accommodate, and everyone needs to use. In the desktop PC heyday, Microsoft owned this “OS” inasmuch as it controlled Windows and Office. Today we have a few contenders for the throne, and what exactly that throne will be built on has not been settled yet. Zoom, of course, takes the video call as the atomic unit or “self organizing criticality” of the modern workplace, and seeks to use that as their foothold to ladder up to other areas. For Slack, it is the employee conversation graph (at rest and in motion), for Okta the theory is that enterprise identity management is how the space will be organized. For Microsoft it is Office 365 itself, including Teams as the “interface.” Facebook lurks around the edges of this fight with Workplace with the theory that an internal social network will become the “hub.” Amazon has WorkDocs, Chime, Alexa and AWS’s secretive “no-code” project, but their play here is, for now, mostly shrouded by the fog of war.
- An immense, epoch defining war for the cloud and what comes after cloud sits as the backdrop to all of thisAmazon and Microsoft are in a trench war at the moment in the west when it comes to cloud (Google is technically in this too but I am not sure that their heart is). Azure seems to be getting some traction off the back of Microsoft’s enterprise bonafides, but AWS is still the elephant in the room. The commodity fight for low level compute and storage rages on as both companies slowly grind out performance and cost improvements and technologies like Kubernetes further commoditize this base layer. But the real conflict is emerging up the stack at various conceptions of what the “platform” and “application layer” means. From the data/ML stack to serverless component ecosystems, the war for the cloud is rapidly leaving the cloud itself behind and focusing far more on what emerges on top.
To understand this move and put it in context we are not just talking about the video call market here, but almost the entire enterprise software and cloud space. Everyone is looking for the “iPhone moment” where the critical bundle of jobs to be done is firmly seized by a transformative player and the rest of the industry organizes itself around that player.
Zoom’s horse sized duck and Amazon’s duck sized horses
Zoom’s primary strategy (at least what we popularly think of as their strategy) has been pretty simple: make the best video conferencing app in the world. To be sure, their success along this dimension has been undeniable. For a category with many incarnations – WebEx, BlueJeans, Hangouts etc – and one which is definitely not new, Zoom has done incredibly well even before the pandemic with its promise of video calling that “just works.” It has been this focus on video calling as “the product” that has gotten Zoom to where it is. You might think of this as Zoom going all in on its “horse-sized duck” – the Zoom application. Zoom’s horse-sized duck has drawn serious blood from its direct competitors in the video calling space (BlueJeans, for example, decided to throw in with Verizon recently). Microsoft’s horse-sized duck – Teams – is much more bloated and is deeply integrated with the Office 365 work stack.
Amazon, on the other hand, is attempting to “win” with a thousand “duck-sized horses” strategy. Core to this Slack-AWS partnership is the integration with AWS Chime. Chime is primarily a communications infrastructure product not too dissimilar from what a Twilio might offer, but can also be run as a consumer-facing application (apparently with mixed adoption results across Amazon employees for voice/video calls). That Slack has decided to not just expand its general adoption of AWS (Slack has always primarily been an AWS shop) but specifically to adopt Chime as the new guts of its beleaguered “Slack Calls” feature is significant because of the strategic importance of that particular set of cloud APIs – APIs which are much closer to customer value than, say, AWS Route 53.
In the immediate term, this makes Slack a much more formidable “friendpetitor” to Zoom. While Zoom has long been integrated loosely into Slack for starting calls originating calls inside messages, a viable Slack Calls (which it has not been up until this point) allows Slack customers to get the value of the “whole product” in terms of workplace communication without needing to pull in a third party like Zoom. The benefits of this might range from bundle economics when it comes to inking enterprise agreements, to crowding out a dangerous “friendpetitor” with dreams of ecosystem dominance. After all, Slack wants its platform to be the platform.
AWS has helped Slack beef up its own horse-sized duck to go up against Zoom and Teams. But stepping back once more and looking at this through the lens of Amazon vs. Zoom or Amazon vs. Microsoft, this Slack-AWS partnership shows us that your horse-sized duck needs to be prepared for an onslaught of duck-sized horses. Because, of course, this is not the only use case of Amazon’s Chime platform. Such a big platform consumer as Slack will surely juice Chime’s reputation and Amazon’s investment in the service, but the game definitely does not end here. What happens if AWS is able to convince Dropbox to implement a similar integration with Chime? What about Salesforce or Okta? What about a totally new startup focused on specific vertical remote working experiences for the accountancy industry? Can Zoom’s horse-sized duck take on all of these duck-sized horses built on Chime – large and small? I’m not so sure.
Asymmetric competition, ILC and the end of software moats
For the past few years I have been obsessed with the idea of the evaporation of software as a moat, primarily as a result of “serverless” technologies and “servicefull” ecosystems. The basic theory (which I will elaborate on in a future post) is that the abstraction of infrastructure entirely away from the developer (with services like Lambda and DynamoDB) allows for very interesting “new value” to emerge on top closer to business value. As Simon Wardley lays out in his seminal post “Why the Fuss about Serverless?” one important upshot of this is likely to be the emergence of an ecosystem of high-level components/primitives which developers (and soon, non-developers) will be able to easily snap together to create bespoke experiences easily. I recommend folks read this post and more from Simon to get a sense of what serverless means for IT & strategy.
One very important concept that Simon brings to bear here is his “Industrialise-leverage-commoditize” cycle for platform owners. The basic premise is that platform owners – companies who operate enabling infrastructure with defined interfaces, functionality and billing like compute, storage and payments – are able to play a very sneaky game with the “customers” that build on top of them, using AWS as the seminal example. After “industrializing” an activity/component like compute, AWS has been able to closely monitor what people do with compute and identify areas for fast-following and further investment, not simply by watching what is publicly available information about those customers but by actually keeping track of the consumption metadata of those customers in order to have essentially inside information on future signals and trends. AWS is then able to use that information to help it decide what to build on top of compute to serve even more and more customer needs, potentially eating some of its customers along the way. A similar dynamic plays out with the (delicate) relationship between the App Store ecosystem and Apple’s own first party services. Below is a visual representation from Simon on ILC in action
Source: Simon Wardley
So, returning to the discussion about video conferencing and workplace software. AWS’s current play against Teams, Zoom and the rest of them seems to be that they will arm Slack and use Slack as the tip of their spear (after all, they can capture a very large part of the upside just by selling Slack the cloud infra that they will use to fight the others). This may even bring the two companies closer together for a potential acquisition if the Slack+AWS workplace “stack” starts to gain traction. But what Slack must be extremely careful about here is the risk that AWS uses this as an opportunity to not just bootstrap the Chime arm of the AWS ecosystem, but that AWS will actually be able to use this partnership as a foothold into displacing Slack itself at some future juncture. Because while Chime’s platform today does not allow Amazon or someone else to easily “roll their own slack” in the same way that it lets them “roll their own Zoom,” there is no guarantee that that isn’t where Chime (or some other service) from AWS goes.
For now, Slack appears safe. There are no obvious high-level primitives from AWS that cater for things like chat rooms, threads, direct messages etc. But the nature of asymmetric competition like ILC is that it sneaks up on you – the floor turns to lava underneath your feet. If you squint and look at what is being done with things like AppSync and AWS Amplify, you start to see the components that could threaten Slack, Twilio, Salesforce and any number of “friendpetitors” to AWS emerging quite quickly. With this Slack-AWS partnership, Slack has essentially “kept its enemy close” in Amazon – a move which might pay medium term dividends but carries extreme long term risks in a game against such a sophisticated player as AWS.