Decision Making & Software Valuations
Reframing software valuations as the product of continuous decision-making frameworks and actions.
Join the global UC family and get the best of tech, business and investing, both globally and as it applies to Asia direct to your inbox 💗
Like what you read and think someone else will too? Don’t let them miss out! 😃
As the world transitions ever more towards the information age, I have been struck by how much business success is based upon consistently “good” decision making frameworks within software companies. Amazon is often held up as the prime example of being able to innovate consistently at scale but considering how important software has become, there are many other companies that fit these criteria. At the same time, the outsized role of decision-making frameworks and their subsequent impact has forced me to reassess how to think about equity valuations of software companies, especially as traditional valuation approaches find it harder and harder to capture the value inherent in these companies.
This article attempts to break down the interplay between both of these core issues for software business builders and investors at a high level. I also reference specific company examples that I have come across, with a specific focus on listed companies. In future pieces, I will try and delve more deeply into the various moving parts.
Underlying context:
The world is becoming more driven by pareto principles and “fat tail” outcomes are increasing, particularly as people's professional and personal lives are driven more by the production and consumption of information / experiences rather than goods. Winners, once identified, will keep winning at an accelerating rate. Given it is reasonably difficult to identify these winners upfront, it encourages making a bet across more opportunities (public markets investing with a VC mindset). A public equity portfolio ends up being a series of call options with real options embedded within them.
It is very difficult to succeed over a period of time as a single product company if you’re in the software (or broader tech) industry. Consumer behaviour across both B2C and B2B lends itself to the aggregation of utility from trusted counterparties - time / attention is finite (as is screen real estate) and the onslaught of information is relentless. At the same time, the low / zero marginal cost of software incentivises reducing the price of any given product to gain scale and therefore, often the only way to increase margins as a whole is to expand out to new product lines or sales channels.
Complexity does not increase linearly as the scope of a business expands and more people join an organisation. It tends to increase exponentially. The impact of this from a cost perspective is further amplified (to the power of N) for each layer of potential decision making, due to the direct cost of decision making (time spent) and the indirect cost (opportunity of cost of decisions that could be made).
Why do decisions that people make matter so much?
Software-based products are in a state of ongoing flux. Unlike physical goods, the software ‘supply chain’ that ultimately leads to a useable product can shift on a daily basis as tweaks are made. Unlike service based ‘goods’, a software product, especially in the cloud era, is largely heterogeneously distributed across the customer base so the impact of decisions made in creating or altering it is magnified very quickly. Layered on top of this is a level of inherent uncertainty and experimentation required to create effective software products.
What exactly are some of the main decisions required in the context of developing and selling software?
Determining which metrics matter and what “success” looks like for a new feature or product release across the total lifecycle of the product.
Example: Amazon’s press release strategy where they actively determine how a product or feature will be sold and described before starting to design and make it. Each core element of marketing strategy is determined ahead of time and ratified so it is very clear how a product should be built and what the associated economics of it is upfront with a high degree of confidence.
Determining the best infrastructure for managing information flows within the broader organisation, as it relates to product development, business development and customer success and putting the relevant frameworks to ensure that people stick with it.
Examples: Creating a template for feature development specifications that resides on an internal wiki system that enables people from across the company to get access to this information and provide iterative feedback, which can prevent a need to alter the feature in the future and minimise the number of “bugs” or errors. The nature of the template, how much information is required to be entered, and how easy it is to comment and provide feedback are all variables that determine usability and must be set upfront to ensure effective engagement happens within the organisation.
In the context of B2B enterprise businesses, creating the appropriate frameworks and expectations for proofs of concept, proofs of value and production-grade contracts.
Example: A software vendor aiming to enter into a new relationship with a large enterprise will need to co-create a series of metrics or acceptance criteria for the initial proof of concept, which should lead to a bigger proof of value and finally to production-grade deployment. For an effective outcome, all relevant acceptance criteria and reporting mechanisms need to be set up in advance and then be monitored carefully. This process is repeated multiple times for true B2B enterprise business development.
Creating the right conditions for ongoing discussion, debate and scenario analysis within teams and between teams whilst at the same time maintaining or increasing the cadence of product releases.
Example: Most performant modern software organisations operating at a high scale operate on services orientated software architectural principles. However, this is not just a technology decision. Underpinning this operating paradigm is the cultural rhythm of information and idea exchange. Amazon pioneered the use of microservices in its systems and team design but arguably, operating principles such as having coherently written analysis as the basis of the discussion in each important meeting contribute hugely to a consistently high level of performance at scale.
There are many more but note a common factor within all of these vectors of potential decisions to be made: actions taken are not in isolation; at some level, they all involve a complex interplay between different people, different styles of thinking and operating and across different time frames. Each decision also has a tendency to compound on the decisions made laterally and also through time and the resulting dynamics are the basis of positively reinforcing flywheels that exist.
So how does this play into software company equity valuations?
Because the impact of the underlying quality and cadence of decision making has such an exponential impact on the future outputs of the organisation, I have begun to frame equity valuations in the following way:
1) The current enterprise value of a company reflects the market’s best guess as to the quality of decision making across the portfolio of products or services a company offers based on currently visible outputs (such as current sales, earnings, margins, cashflow conversion etc.).
2) The valuation gap between the current market price and “intrinsic” value then reflects the lack of understanding the market has on the nature of decision making as it relates to:
The ability to create and launch products into the market that do not only open up new revenue streams, but do so in a way that minimises the number of decision points that go into making those products, and ideally allows each product to positively interact with the others (especially if network effects are present).
Putting in place core infrastructure that creates resilience in terms of the ability to adapt quickly to changes in the external environment with the least number of decisions that need to be made.
Maintain or increase product development cadence and customer acquisition efficiency per head as the size of the organisation increases.
Dynamically determining the appropriate type of pricing model / strategy depending on the nature of the operating environment faced by the company.
As these factors become bedded down and their impact becomes more visible in reality, we tend to see revenue acceleration and concurrent sales efficiency gains, which given software margins are so high, leads to financial results well ahead of typical consensus expectations and a repricing upwards of the present value of future cashflows flowing from the company. This effect more than compensates for any macroeconomic changes, such as long term nominal rates that can negatively impact the assessment of terminal value in these high growth companies.
Furthermore, companies that exhibit this type of behaviour end up being more reflexive from a financing perspective too - they find it easier to raise new capital at attractive valuations for growth, which further leverages their decision-making capabilities.
Here are some of the best examples that I have personally come across that show these dynamics in action:
1) Twilio (TWLO): Years of investment in putting Amazon-like team and operating infrastructure in place, as well as building a strong base of grass-roots developer support along with some brilliant acquisitions, has led to the rapid evolution of the company towards a full-scale user engagement journey data management platform. Not only does the product set allow for communications across multiple channels to be managed but it allows for complete tracking of customer engagement, essentially supplanting the role of a traditional CRM as the source of truth of future business activity. The latest set of results bear this out with revenue blowing past consensus expectations, guidance being revised upwards and dollar-based net retention rates (DBNER) increasing to c140%, which are best in class for any software company.
2) Cloudflare (NET): At the latest investor day, Cloudflare’s management specifically spoke about the rate of product development increasing further in 2021 and beyond. Cloudflare has been building new product ranges on top of its own development infrastructure further shortening time to market and validating its own infrastructure platform (Workers) for others to build on top of. At the same time, DBNER has been increasing and the QoQ sequential growth rate of large enterprise wins is larger than the YoY rate of large enterprise wins of its nearest competitor (Fastly). As new sales teams bed down, sales velocity should increase even further.
3) Twitter (TWTR): Years of core investment in the underlying platform, especially around rebuilding the core stack under a microservices model now means that product development velocity is increasing. It is not a surprise that Twitter is executing future revenue-enhancing product acquisitions like Revue and introducing Spaces at precisely this point, as well as making a conscious decision to transition towards more subscription-based revenue streams. User growth and ARPU now seems to be meaningfully inflecting upwards.
4) Peleton (PTON): Prioritisation of the development of social features has led to people buying multiple devices to participate in the community. At the same time, engaging content keeps them coming back for more. Peloton can essentially choose whether or not to engage in product marketing making their CAC completely under their control whilst at the same time improving their revenue base through increasing sales channels and hardware products.
5) PayPal (PYPL): Similarly to Twitter, decisions made in terms of moving towards services-based architecture and conscious decisions made to create a multi-layered platform (payments, BNPL, crypto, moving into stock trading), as well as owning more of the customer engagement journey in e-commerce (through their Honey acquisition) are starting to pay off in terms of user numbers and revenues inflecting upwards. Their recent analyst day made it clear that decisions being made are focused on optimising around these network effects with clear thought gone into how customer personas can interact with multiple products over time (e.g. customers using BNPL to transact more efficiently can eventually utilise some of their personal working capital to buy into crypto).
6) Mohawk (MWK): Investments made in building out marketing and data analytics infrastructure are now allowing Mohawk to easily identify brand categories that people actually want to buy and buy out emerging vendors in those categories at attractive valuations. Importantly, cost synergies can actually be realised easily and as their scale grows, further acquisitions become easier to source, execute and integrate.
7) Elastic (ESTC): Years of building the core technology are now allowing Elastic to more easily build new products (logging, monitoring, security) on top of the underlying technology stack at speed and scale. A conscious decision to stay true to their open source roots also allows feature enhancements to come through at speed and scale. At the same time, as the sales organisation continues to get bedded down, revenues should continue to inflect upwards and sales efficiency should continue improving.
The list can go on and I’ve missed some notable examples of this by Snowflake but I really just wanted to focus on the ones I currently own.