The New Software Stack

The New Software Layered Cake

Part 1: Uncovering Technologies Underlying Today's Digital Economy

Sekhar Sarukkai

Over the last decade, we have seen a tectonic shift in the software stack, unlocking many opportunities for entrepreneurs and investors alike. This trend continues with Cloud-native applications and data-centric applications that are poised to change the landscape of companies for the next decade.

In this 2-part blog series I will explore my own understanding of the new software ‘layered cake’, who is eating this cake today, and mapping my own angel investments in early stage companies to this framework. I look forward to your valuable input as I refine my views on the rapidly evolving landscape. It is possible that I have over-simplified or hyper-extended on specific instances, but I hope that this framework helps some of you place the various opportunities you may be evaluating, in context.

The new Software Cake

The new software cake is vegan. It is not the cake of yesterday - no eggs, no butter, and if gluten free, no wheat either. It is something you wouldn’t have recognized as a cake a few years ago.

This new software cake is a Service, continuously developed, deployed and hosted in the Cloud, built with significantly lesser number of developers (https://www.morganstanley.com/ideas/new-software-stack-2019), built with >90% of the software relying on open source software or with no-code at all, and continuously deployed with self-describing Infrastructure-as-Code templates, built on, or as, SaaS APIs, or data-centric applications. In contrast, traditional software stacks were thought of as a three-tier web application deployed once a quarter and with rigid interfaces and fragile network controls.

So what?

When talking of a software stack, many practitioners will list out LAMP (Linux, Apache, MySQL, PHP), or MEAN (MongoDB, Express, Angular, Node.js). An operations team member may list the Apache Cloudstack that is used to deploy and manage large virtualized applications. However, today’s software is not typified by a standalone application or silo’d stacks, but by an integrated, continuously deployed, and observed Service that in turn incorporates and depends on many other Services to achieve its desired functionality. I contend that a true view of the modern software stack can be obtained only with this Service perspective, and hence this article tries to portray this larger software stack. The goal is not to minimize or downplay the importance of the application stack but to shed some light on the broader software stack for readers interested in framing the various companies and Services that form the software industry today. There have already been trillion dollar opportunities across various layers of this new stack over the last decade, with a similar amount of opportunity to be unlocked in the next.

Let us look at each layer in the above "new stack" in a bit more detail:

Layer 1

Layer 1 is the platform on which the bulk of the software is deployed. This layer is dominated by trillion dollar companies offering public Cloud platforms including Amazon AWS, Google Cloud Platform and Microsoft Azure. In addition to the Cloud, there are innumerable open source platforms such as open source Linux & its variants, Kubernetes and containerization, and language compilers and runtimes on which today’s software is built in public or private Clouds.

Layer 2

Layer 2 contains two categories of technologies: App infrastructure and Data infrastructure collectively accounting for a trillion dollar market value today. This layer forms the foundation of modern apps that are developed in a Cloud native form and is built on top of Layer-1 technologies.

App infrastructure covers the infrastructure needed to simplify development with low-code/no-code platforms (ex. Airtable, Unquork), continuous software development (ex. Git), test (ex. Postman) and deployment tools for creating a software factory (ex. Gitlab, Jfrog), automation (ex. Hashicorp), and feature controlled product deployments (ex. Split) to ensure Service issues are identified and resolved in a continuous basis with granular control of every customer’s unique experience.

In addition to the app and data infrastructure, this layer also has core monitoring/observability products (ex. Datadog) and security infrastructure (ex. Okta, Skyhigh/McAfee) required for continuous monitoring and security of data and applications.

Layer 3

Leveraging the app and data infrastructure, this layer enables headless Services delivered as APIs that perform some core business function, and Services that aggregate disparate data either near the source (edge) or in the Cloud for further upstream value creation.

SaaS APIs are a class of services that are vertical or horizontal API based headless SaaS services. These Services offer a best-in-class core functionality typically powering other businesses or SaaS offerings and are not typically an end-user business SaaS. There is more than 100B+$ in valuation of companies in this layer. Twilio and Agora are two examples of collaboration APIs, while other popular APIs include large Services like Stripe (payment). There are many new entrants in this category as well, such as specialists Auth0 (authentication) and Scale (AI data API).

Integrated Data Services are Services that integrate disparate data (internal and/or external to an enterprise) to unlock value. These companies typically specialize in a vertical. Examples include Plaid (financial data), Hubspot (marketing data), Segment (customer data). There are multiple opportunities in this category for newer companies as well; some examples of early-stage companies include Noyo(insurance data), Particle(healthcare data), and Argyle(employment data).

Edge computing companies focus on performing computing closer to the source of data in a distributed manner. The advent of 5G and a move towards remote-native workforce makes this move to edge computing more practical and imminent. This move has the potential to reshape many offerings in different layers of the stack over the next couple of years. A majority of investments in this category focus on autonomous vehicles such as Meuro and Aurora, but increasingly is finding use cases in verticalized applications such as factory floor automation and other IoT related use cases as well.

Layer 4

The next layer in the stack includes some of the largest vertical and horizontal SaaS vendors (usually adopting lower layers of the stack) to rapidly develop and scale their own applications, and accounts for more than a few trillion Dollars in market-cap. This layer can also include traditional applications deployed on premises but these will increasingly leverage all the technologies and processes used for a public SaaS, hence these apps can essentially be considered to be private SaaS applications.

Horizontal SaaS companies include IT services that are relevant across any industry vertical and include collaboration services such as Zoom, Microsoft O365 and Slack. Business SaaS including services like Salesforce and Workday that address core functions in any enterprise (such as sales and HR) also falls in this category.

Vertical SaaS include Services that target specific industry verticals such as Veeva(healthcare) and nCino (fintech) built on horizontal SaaS platforms. While these Services are focused on a niche vertical solving a real domain specific pain point, there are other vertical microSaaS that are Services addressing a niche problem such as the thousands of microServices in the Shopify app marketplace that enable e-commerce sites to help with shipping, payment, billing, etc. Other vertical SaaS includes companies such as Noodle.AI that helps enterprises and manufacturing companies reduce waste across complex supply chains.

B2C SaaS includes many diverse applications including personal fin-tech (ex. Robinhood), wellness applications (ex. Headspace), insurance (ex. Lemon). This broad category also includes startup unicorn companies like Carta that help manage employee stock grants/equity, and Oscar health that provides more control of health data to consumers.

Layer 5

This layer essentially is an aggregation and orchestration layer across Services, APIs and data that have been created in lower layers of the stack.

Business process automation(BPA) is a holy grail that has long fallen short of revolutionizing both robotic workflows as well as automating complex processes that can enable an agile business. More recently, robotic-process-automation products (ex. UIPath, AutomationAnywhere) have made strides in reducing robotic manual work in enterprises. With the advent of modular building blocks and APIs in the lower layers of the new stack, it is now feasible to actually put together a robust business process such as quote-to-cash, order-to-bill, etc. BPA as a Service can now meet the needs of a modern enterprise with orchestration tools (ex. Zapier) that allow plug-and-play of thousands of Services discovered in marketplaces, and consumed as horizontal or vertical SaaS (or microSaaS) with limited friction. There are many young startups in this category addressing varied vertical markets (ex. Shippo for logistics, Hyperscience for Finance & accounting, and Tekion for automobile dealer management systems).

The final category in this layer is Marketplaces that include a broad swath of companies from gig economy companies such as Uber and instacart that connect gig workers with activities (such as delivery), to digital currency marketplaces such as coinbase. In general. Marketplaces address a vertical industry by enabling the supply or demand side. Shopify is an example of a marketplace for easy creation of e-commerce sites enabled by a rich set of third-party vertical microSaaS applications that fulfill key business functions. Other marketplaces such as Roblox are a marketplace for developers to create new games that can be monetized on their platform.

Summary

The new software stack is already forming and is the engine behind the digital revolution today and into the foreseeable future. In part 2, I will examine characteristics of example early-stage, angel funded, startup companies mapped to different layers on this stack.

Also read: New Stack Brings New Opportunities