Software Primer, what Everyone Must Know

Welcome to the 104th Pari Passu Newsletter,

Today, I have a very not-contrarian topic for you: a primer on the software industry. While the software model is generally considered a high-quality business, recent times have shown us that things can go bad quickly when you put 10x leverage at S+800 on a great company.

Therefore, even restructuring professionals need to be knowledgeable in this sector. After all, software is undoubtedly one of the most important industries, impacting nearly every aspect of the global economy. From the software that powers our phones and laptops to the software that helps NASA launch and operate its rockets, software is ubiquitous. Much of the rapid innovation and technological advancement of human civilization in the past few decades can be attributed to the growth of the software industry.

Let’s get to it.

Invest in alternative assets inside your retirement account with Carry

Carry enables you to invest in startups, real estate, and other alternative assets inside your Solo 401k or IRA.

With Carry.com you’ll be able to access:

• Tax-free growth within your IRAs or Solo 401k

• The ability to invest in alternative and traditional assets all in one platform

• Strategies like the Backdoor Roth and Mega Backdoor Roth conversions if you’re beyond the income limits

With Carry, your alternatives can grow tax free. Use the code PARIPASSUBF for 50% off your first bill on the Basic and Pro plans, regardless of monthly or annual — pick the best reward for you and your alternative investments. (Expires November 30th)

Overview

The software industry comprises firms that employ software developers to write code in order to develop software with a variety of business applications. The firms then distribute their software through a variety of different channels. Software firms have no raw materials, packaging costs, or inventory as their product is entirely digital [4].

Structure

There are two main types of software that firms develop: application software and infrastructure software. Application software performs a specific task that is unrelated to supporting other software. Infrastructure software provides an environment for other software applications to run and interact cohesively. Examples of application software include productivity apps such as Microsoft PowerPoint and Excel, customer relationship management software (CRM) such as Salesforce Automation, and enterprise resource planning platforms such as SAP ERP. Examples of infrastructure software include security systems such as firewalls, network software such as Cisco's Internetworking Operating System, and operating systems such as Microsoft Windows [5].

Market Share

Relative to other industries, the software industry is quite diversified. Leading the pack is Microsoft with 26% of revenue market share, significantly higher than second place Oracle with 6% and third place SAP with 4%, followed by Salesforce and Adobe. In fact, there are 72 public software companies with market caps greater than $5bn, a testament to the diversity of the industry [5] [11].

Software is characterized by low barriers to entry, so the formation of new businesses is high. This is in large part due to the versatility and reproducibility of software, meaning there are effectively infinite software packages for infinite applications that theoretically anyone with a computer can produce. These factors make it difficult for a few firms to capture a majority of the market [5].

Firm Strategy

Product Delivery

Software firms have delivered their products through a variety of different channels. Here is a list of the most common methods [2] [5]:

• Perpetual license: clients are sold a perpetual license to the software, incurring a one-time fee; updates are infrequent

• Term license: clients are sold a license for a specific period of time; there are no updates

• Subscription: clients “subscribe” to the software, much like a streaming service subscription; updates occur regularly

• SaaS: similar to subscription delivery but the software is delivered via the cloud as opposed to residing on customer premises; updates occur frequently

• Open source: software freely accessible by anyone on the Internet

Companies will typically offer multiple delivery options for their software, such as selling both perpetual licenses and SaaS delivery. This provides their clients with the flexibility to choose the method of delivery that best suits their needs. The distinction between purchasing a recurring subscription versus a perpetual license is much like that of deciding between renting or buying [5].

Why does everyone love Software as a Service (SaaS)

The SaaS model was first introduced in 1999, when Salesforce launched its customer relationship management (CRM) platform. This was the first SaaS solution ever developed, and seeing that Salesforce is now the fourth-largest software company by market cap, it was a revolutionary advancement. Ever since, SaaS has become the dominant delivery model in the industry, for the huge upside that it provides to both the vendors and customers of software [9].

From a vendor perspective, the SaaS model provides a high degree of revenue visibility due to the recurring nature of subscription-based revenues. Additionally, vendors only have to maintain a single code base, of the software’s latest version. Any new updates can be rapidly introduced in just days or weeks as the software is entirely cloud-based. These factors help companies maintain lower upfront costs, allowing for rapid market expansion [5].

From a customer perspective, an SaaS model also helps reduce upfront costs by replacing a large, one-time cost with smaller, recurring subscription fees. The software can very easily be set up and running since everything is via the cloud, and customers always have the latest version due to rapid updates [5].

However, the SaaS model is not flawless. For vendors, delivering their software products via the cloud means that they have to support higher initial costs for software hosting. This is not a concern with most other delivery methods, in which the software is hosted on customer premises. Additionally, subscription-based revenues means slower revenue recognition. For customers, dependence on the cloud poses greater data vulnerability and security risks [5].

Capital Allocation

Company maturity dictates capital allocation. In early-stage software firms that are still pre-revenue, product development is the main priority. These types of firms invest heavily in R&D in order to rapidly develop a market-ready product. Newer companies that have a marketable product will shift their capital towards sales and marketing (S&M) spending, in order to create a sustainable and growing client base. S&M spending is a major contributor to the high CACs (customer acquisition costs) that software companies typically incur [5].

As a software company matures through securing a reliable client base and beginning to generate steady cash flows, it may begin allocating capital towards share buybacks and M&A. Share buybacks are a strategy employed by firms, not just in software, as a way to elevate investors’ returns and boost performance metrics such as earnings per share. M&A, specifically in the form of acquisitions, is also common for larger software companies. These are typically strategic purchases, used to expand product scope and capability. However, software deal value and volume have fallen off in recent years due to less clear macroeconomic conditions [5] [6].

Software companies cultivate growth via different strategies. Newer firms depend on organic growth, such as through product development and benefiting from open source contributions, while established firms depend on M&A and various market strategies in order to pull for growth [5].

In 2023, according to PitchBook data, the median R&D allocation for software companies was 24%. The median allocation on share repurchases was 9% and on M&A was 1%. However, large firms such as Oracle, Shopify, and Snowflake had much larger M&A spend, with the firms allocating 55%, 31%, and 18% of revenue respectively to acquisitions. The median dividend allocation was 0% as very few software companies offer dividends [5]. This would theoretically imply that software companies see attractive internal capital allocation opportunities (or are too afraid to admit they have no use for incremental capital).

Financials

Tailwinds and Performance

Software’s growth can be attributed to a few key factors [10]:

• Increased Internet access: in 2007, just 7% of the world had Internet access; today, over half the world has Internet access. Similarly, half of Americans had internet access in 2000, but today, over 90% have access. This development created a rapidly growing market for software to cater to.

• Development of the personal technology industry: unlocking one of software’s largest markets, the consumer, was dependent on the development of the personal tech industry. The advancement of cellphones, laptops, and other personal devices by companies such as Apple, Blackberry, Dell helped proliferate tech, and naturally software, into the homes of millions of consumers.

• Improved business operational efficiency: early software also catered towards business applications, and early adopters of software often saw significant improvements in operational efficiency and thus overall performance. As the world became increasingly “online”, software integration quickly became a necessity for any competitive business.

Over the past year, the industry is up an average of 23.7% with an average annual earnings growth rate of 19% over the past three years. The average P/E ratio of software companies is currently 49.3x, significantly lower than the three-year average of 61.0x. Experts forecast the industry will grow by an average revenue growth rate of 10+% through 2027 [1] [4] [5].

Software has historically remained quite shielded from global economic downturns, and actually often expands during recessionary periods. During the 2008 financial crisis, software growth fell, from a high of 9% to a low of 3%. However, U.S. GDP growth saw red, steadily declining from a growth rate of 7% in 2006 to a low of -2% in early 2009. During the 2020 pandemic, U.S. GDP growth once again went negative to around -1%, but the software industry actually experienced significant growth due to the mass transition and dependence on the digital world, plateauing at 7% in 2019 before climbing to 13% by 2022 [5].

Financial Statements and Metrics

Financial statements of software companies contain their own nuances and distinctions that make them unique from those of companies in other industries. While these trends are not present in all software companies, they are important to keep in mind when analyzing these types of companies [5].

• Revenue: most firms utilize subscription-based models, generally entails receiving full year payment upfront and then gradually recognizing the revenue over the given period (more on this when talking about deferred revenue)

• COGS: less important due to the lack of packaging, inventory costs; mainly represents hosting costs for maintaining and staffing data centers or outsourcing the labor to cloud providers; typically COGS make up less than 30% leading to 70%+ Gross Profit margins

• Research and Development (R&D): high in early-stage companies during product development and still sustained at maturity given this is where all development costs (i.e., software engineers salaries, one-off infrastructure costs, IP and patent costs) get reflected in. Worth noting that R&D expenses can show up both on the Income Statement (as Operating Expenses) or in the Cash Flow (as a form of CapEx)

• SG&A: marketing expense are a key driver of software unit economics (we will touch later on the LTV to CAC metric) and are a major cost bucket which often makes up 30%+ of revenue

• Adjustments to derive adjusted net income: many firms use stock-based compensation (SBC) to attract and retain employees; on this topic, investor love debating as many argue that SBC should be added back as an adjustment given it is not a cash expenses, but it is essential to remember the resulting dilution associated with this expense (said simply, if you add back SBC but do not include incremental dilution in your valuation, you’re implying that this expense does not exist which would lead to massive mistakes in valuation given how significant this cost can be for software companies)

• Deferred revenue: an important line item due to how revenue is accumulated by software firms; to understand this line item, let’s look at an example: a customer pays an annual subscription on December 31st, the company will record the entire balance as deferred revenue in that moment, and every quarter the deferred balance will decrease combined with a quarter of revenue - and costs - recognition; this item is key to understanding why software companies have often negative working capital which explains how growth can have a very accretive cash impact (when new customers pay for the service, the entire cash balance is received but no revenue / costs are recognized)

Alongside the common valuation and performance metrics that analysts use, software companies are also evaluated based on a set of industry-specific criteria [5]. 

• Bookings: total value of all software contracts signed within a given period; measures how much new business is generated during a specific period

• Revenue performance obligation (RPO): also known as “backlog”, measures the revenue that a company has contracted to earn in the future; an important forward-looking indicator that measures how much future business has already been confirmed

• Annual recurring revenue (ARR) measures the amount of revenue that is recurring; an important aspect of software firms that provides steady future cash flows and compensates for the high initial R&D spend and customer acquisition costs (CAC)

Economic Moats

Switching Costs

Competitive moats in software are typically centered around switching costs. Switching costs are drivers of high customer retention, which is key for subscription-based models. Economic moats in the industry are disproportionately based on switching costs, as other common moats such as intangible assets and economies of scale simply do not apply [5] [7].

Changing software applications is often a costly and time-consuming process, from both a monetary and human capital perspective. Once a software has become integrated with other applications in the organization and employees have become proficient with it, switching to a new software would impede operational efficiency [4] [5]. 

Per a 2023 MuleSoft and Deloitte report, companies use an average of 976 software applications. A high-level CRM system, such as Salesforce, integrates many of these applications, combining the company’s marketing, e-commerce, inventory management, customer service, and much more into one comprehensive system. Attempting to replace this system would be extremely difficult, risking significant data loss, disruptions to operations, and reduced employee productivity [5].

In this way, switching costs make customers less likely to begin buying from a competitor. Two key metrics that software firms use to measure the strength of their switching costs is gross account retention and net dollar retention. Gross account retention measures the percentage of clients who remain customers; for example, a 90% gross account retention means out of a cohort of ten clients, nine remain clients after a year.  Net dollar retention measures how much the company’s annual recurring revenue has grown or shrunk versus an earlier period. Maintaining high gross account and net dollar retention rates is important to offset the high upfront spend many software firms incur from R&D, S&M, and CAC. The industry average gross account retention is ~93% and the average net dollar retention is ~115%. Naturally, firms with stronger economic moats enjoy higher retention rates [5].

Network Effects

Network effects play a backseat role in software and are nonessential for a company to succeed. However, most mature companies benefit from some type of network effect that ultimately strengthens their economic moat.

Network effects are important in developing ecosystems around software companies – in other words, network effects help create environments that consist of more than just a buyer and a seller. As a software scales in popularity, it attracts third-party developers to create plug-ins or supportive applications. For example, in video games, these supportive applications might come in the form of modifications, “mods”. Popular video games such as Minecraft and Grand Theft Auto 5 have thousands of mods and highly active modding communities, which both improve gameplay quality and content. A relevant professional example would be Microsoft Excel, with tools such as SAP’s Analysis for Office Excel improving the software’s data analytics capabilities [5].

Over time, well-developed add-ons to existing software attract increasing amounts of users to the platform. This creates greater demand for additional software functionality, attracting even more external developers. This positive feedback loop not only grows a company’s client base, it also strengthens its existing switching costs. Now, not only would a customer have to be willing to give up the base software’s capabilities and their familiarity with it, but they would also have to relinquish a community of developers looking to actively improve the software [5].

High Upfront Investments Support Moats

Software spending is often front-loaded. These high initial costs are aimed at constructing strong economic moats, in the form of switching costs and network effects. Eventually, these moats help the company generate years of “sticky” revenue through strong customer retention. Software firms strive to maximize the lifetime value (LTV) of their customers – the industry median LTV to CAC ratio, which compares the total value a customer provides to the company to the cost of acquiring that customer, is 3.5x. An average LTV to CAC ratio of 3.0+ is often considered healthy. High LTV to CAC ratios are positively correlated with stronger economic moats, with wider moats typically producing better ratios [5].

These upfront investments also support high ROICs as a company matures. Wide moats and diversified software portfolios from years of high R&D and S&M spending enable mature software companies to boast ROICs that well exceed the industry average WACC of 8-9%. For example, Microsoft boasts a nearly 50% five-year average historical ROIC. Once again, high ROICs are generally positively correlated with wider economic moats [5].

Current State of Software

Software, primarily SaaS, has suffered slowed growth in recent years. The market cap for public SaaS companies peaked in 2021 at $1.7tn, and has since fallen to just under $1.3tn as of June 2024. Additionally, public SaaS companies have generated negative real returns since 2019, down 20% in the past four years and down 53% since November 2021 as of this summer [11].

Part of this slowdown in public growth can be attributed to the lack of SaaS IPOs. Since 2022, there have been just two IPOs of significant private software companies (Klaviyo and Rubrik). In 2021 alone, there were twenty-three such IPOs. This period of IPO reluctance is not because private SaaS are any worse than their predecessors, as analysts observe, but due to mismatched expectations in the IPO market. Management teams are choosing to be more cautious and thus opting to remain private [11].

Furthermore, incremental (recurring) revenues, SaaS’s bread and butter, have already peaked and have been declining since 2021, from $21.6bn to $18.2bn. Simultaneously, S&M spending over the same period rose from $28.5bn to $37.2bn. Net dollar retention has also peaked and is declining, from a high in Q4 of 2018 at 125% to a six-year lower of 110% in Q1 of 2024. Public SaaS companies are spending more than ever for weaker and less consistent revenue streams. For an industry built upon the recurring revenue model, these numbers are alarming [11].

It is difficult to pinpoint the exact cause of software’s present struggles. Some analysts hypothesize that businesses are still digesting software overspending during the pandemic which helped fuel the industry’s rapid growth in 2019 and 2020. Another explanation is that software is simply maturing and is no longer the high-growth industry that it once was ten years ago. Naturally, margins will shrink while innovation becomes increasingly difficult in a saturated market [11].

Outlook

Tailwinds

Despite recent struggles, software should still be a sector of reliable growth for the foreseeable future thanks to a new development vector in the form of AI. With analysts projecting 10%+ annual revenue growth rates through 2027, here are the key tailwinds supporting the industry [4] [11]:

• Artificial intelligence (AI): generative AI demand is through the roof and will fuel growth, and businesses agree – Avenir survey data stated that 28% of companies believed AI “will be transformational and is still underappreciated”, and 40% of companies believed AI “will help [them] significantly”

• Data: exponential growth in data and the need for storage will only continue to grow; software firms provide solutions to organize, protect, and use this data which originally comes in very raw form

• Public cloud: public cloud adoption is abundant, and many more firms are making the transition; software firms will be the ones supporting cloud integration and transition

• Digital transformation: push for firms to implement digital workflows with many businesses adopting hybrid or fully-remote schedules post-pandemic

Headwinds

Here are the primary headwinds that may impede the software industry over the next few years [4]:

• Higher interest rates: higher rates discourage investment, which is key for many software startups; without the upfront capital, it is difficult for early-stage firms to develop their product and acquire customers

• Economic downturn: customers typically tighten their IT budgets during downturns; this also discourages investment, including in software firms

• Regulation: regulatory intrusion inhibits innovation; a relevant example is the ongoing discussion of federal regulation of AI

Challenges (Data Security)

By far, the most pressing issue that the software industry faces is that of data privacy. Data breaches and cybersecurity attacks are common, even at big-name firms, and these strikes will only continue to increase and innovate over time. [5]

In essence, a cyberattack is an intentional effort to steal, expose, destroy or negatively affect data, applications, and other digital assets through unauthorized access. The average cost of data breaches is $4.35mm, plus the additional unquantifiable damages caused to customers. Cybersecurity breaches can also lead to regulatory fines and legal action taken against the company for loss of personal customer data. One estimate projects cybercrime to cost the global economy $10.5tn annually by 2025 [8].

Businesses depend on security infrastructure software such as firewalls, data loss prevention tools, and antivirus software in order to protect themselves from such attacks. However, hackers are often able to find a way through, thus posing a constant threat to business while simultaneously creating a sustained demand for a cybersecurity market. At the end of the day, cybersecurity breaches are simply an evil that all businesses must learn to cope with [8].

Key Takeaways

In wrapping up this software primer, let’s recap some of our key takeaways from this ever-important industry.

• Highly diversified industry with low barriers to entry: Aside from Microsoft’s market share dominance, the software industry is highly diversified with many firms remaining relevant by filling a niche. The low barriers to entry into software enable startups and early-stage companies to compete and disrupt, even among established giants such as Microsoft and Oracle.

• Switching costs are key to company success: Switching costs drive customer retention, which, in the modern market, is everything for software firms. Firms must be able to remain relevant by ensuring that their customers cannot replace their software – essentially creating an inelastic demand for their product. Constructing an impregnable economic moat through the development of switching costs should be a priority of emerging software companies.

• Cybersecurity remains an area of concern: There always exists the looming threat of cyberattacks. While new software is constantly being developed to improve businesses’ abilities to detect, prevent, and recover from such attacks, hackers are also constantly innovating. There simply is no feasible solution in sight.

Thank you again to Carry for sponsoring this deep dive and for helping us save on taxes, let’s go enjoy some Thanksgiving leftovers!

List of Sources

📚 Interested in our updated reading / wellness list? Check it out here.

📈 Interested in our IB / PE / HF course recommendation? Check it out here.

👕 Interested in our merch store to shop our latest swag? Check it out here.