Mastering Software Product Management: Practical Solutions and Expert Tips for Strategy, Decision-Making, and Building & Selling Market-Leading Software Products

CHAPTER 1

The Practice of Product Management

Introduction

As seasoned senior management of the organization, you have convinced the investors to embrace digital transformations and are now expanding your team with software professionals. Experienced people have advised you to engage product managers to review your organization’s processes to align your digital transformation initiatives with your business goals. An MBA graduate with limited industry experience is hired as a software product manager and is now running from department to department to understand her assignment. Everything she is learning now does not seem to connect to her studies in a two-year MBA program. As a product management leader, you have developed and marketed several products and led product management teams, yet your expertise is only getting highlighted in your functional domain. A cross-functional scope of product management experience is not apparent. If you have dealt with any of these in your career engagements, we try to address some of them in this book.

Structure

In this chapter, we will cover the following topics:

Software Product

Product Management Process

The Framework

Software Product

Every organization delivering software solutions is providing software as a service. Yet, they all talk about software product development and want their development initiatives spearheaded by a product manager. It is the perfect fallacy that every software product manager lives through. Software product managers are engaged in understanding customer requirements. Thus, every product manager has sometimes been questioned why a specific feature requested by a customer could not be delivered exactly as desired. Are business analysts not reviewing customer requirements and helping develop solutions for several decades? Why should the product management function be any different?

As we answer these questions, let us understand what a software product is, why, and what it achieves in the first place. When computers came into existence, they added tools for operating the computer. The computer as a hardware box anyway was irrelevant without the software. Initial computing devices, such as the mainframes, were expensive, and a model would hardly sell in hundreds or thousands of numbers. Suppose a bank needs a mainframe computer for banking operation optimization. It will start negotiating with one of the mainframe providers, such as IBM for the hardware. It will engage Business Analysts (BA) and System Analysts (SA) who will work with IBM on the suitability of the mainframe for the bank’s needs. IBM will provide all the tools for the BAs and SAs to outline the specifications for the system. Engineers from IBM or the bank receive training on the system and tools. They will develop software based on the specifications created by the BAs and SAs. Every system was a closed unit with its custom software. The software was a service for a specific hardware. Only an affiliate of IBM or a bank’s contractor can provide services in such an environment. One has to develop software specifically for every customer. Being a human-intensive activity, you incur the cost for every customer added.

The advent of Unix was a paradigm shift in computing. A high-level language, C, was used to write the operating system. One can now port code from one system to another system. Unix started running on thousands of mini-computers and workstation class machines. It was not practical for the system providers to support all the custom software development themselves. Workstations started shipping office productivity tools such as word processors, spreadsheets, and presentation tools. With many deployments and varied customer needs, the vendors could no longer meet the needs of all customers. They needed to support the developer ecosystem or community. They also licensed software from some of these third-party developers and included it in their hardware. IBM went a step further. They standardized the hardware specification for the IBM Personal Computer (PC). The vendors who sell PCs for these standards can now bundle the OS from Microsoft. With this PC revolution, Microsoft became the largest software-only vendor. PCs were deployed in millions in several households. Unlike workstations in offices, PCs serve purposes beyond productivity tools. They became tools for education, entertainment, programming, and so on. We are talking about at least a hundred tier-A hardware vendors, tens of thousands of software vendors, and millions of computer users and customers; we have left the closed ecosystem of vendors and customers and moved to an open market. Today, even the mainframe vendors provide open systems tools for developing software for the mainframes. For any vendor to understand the whole market is unthinkable. However, software services businesses and custom software for specific customers still exist and continue to flourish.

Economics

When developing custom software for your customers as a service, every customer is a unit with its manpower.

Revenue (R) = R1 + R2 + … + Rn

Cost (C) = C1 + C2 + … + Cn

Ri and Ci are the revenue and cost, respectively, for the ith customer. Every new customer acquisition incurs additional costs. While the profits are there, they are only nominal. Let us talk about Microsoft. Microsoft sells Windows Home editions for under 100 USD. Can a custom OS be built at that price? It is possible as Microsoft is developing the Windows OS only once and making copies for the customers to deploy. Microsoft has n=500 million active Windows devices. Now, the revenues and costs look as follows:

Revenue (R) = n*Price

Cost (C) = C

The Cost (C) gets spread over many customers, and thus the price can be lowered drastically. We saw that innovative software products can reach millions of customers and users. Software products companies can grow big and reach billions of dollars in revenues. Similarly, software services companies can reach billions of dollars in revenues. But they target large customers, generating hundreds of millions in revenue from each. We face two challenges with software products:

How do software product companies ensure that millions, or even billions, of users will use their products?

How do companies spread the costs across the customer base?

When economies of scale are reached for software development, a software product mindset is in the organization; the costs can be spread across customers. It also means no two customers perceive the same value for the same product they use. For example, Microsoft Word for a lawyer drafting contracts or an author writing a novel is economically more valuable than a householder who types a letter or two sporadically. These create a consistent challenge. Companies are building software solutions for a class of customers, where each customer is seeing differentiated values in the same product. Companies are innovatively spreading their costs across customers in different channels. We will see some of the channels.

Sales and Marketing

Are Microsoft Windows and Office not supposed to be product suites or collections of products? It was the case several years back, not today. Microsoft acquired a few market leaders in the fields of document writing (Word), spreadsheets (Excel), and presentations (PowerPoint). Microsoft would sell Word to the company’s secretarial staff, writing official communications. They would type the content in Word, print it, and post it on the notice boards for employee communications. Accounting executives use Excel, and marketing executives use PowerPoint. They were selling three products to the same company in limited numbers. Microsoft has a unique scheme of combining all these products as a suite under Microsoft Office. Office digitally enables a knowledge worker to communicate effectively and should be available on every computer in an office. MS Office is a knowledge worker productivity tool with features of document writing (Word), spreadsheets (Excel), presentations (PowerPoint), databases (Access), email (Outlook), and so on. While Microsoft could only sell ten copies of the individual products, they can sell a hundred copies of the suite. The suite can be priced competitively to the component products, as Microsoft can make more money selling larger volumes. Companies follow such strategies even today.

MS Outlook had two options: MS Outlook shipped as part of MS Office, and MS Outlook Express shipped for free with MS Windows. Today, this free version is Windows Mail and is part of the Windows components. MS Windows as an operating system is no longer just the software to boot a computer. Today, people look at an operating system to provide basic functionalities alongside making the hardware usable. For example, with the operating system, you can connect to the internet and your internet email sites, such as Gmail, Outlook.com, Yahoo.com, and so on. Windows Mail helps you achieve that and is part of the MS Windows product. MS Office Outlook provides connectivity to enterprise email systems and caters to the needs of a different group of users. Microsoft Office Home and Student Edition does not ship with MS Office Outlook. The choice of product bundles in various editions plays a crucial role and can decide the price of a product. We will touch upon this when we discuss pricing. The definition of a product is dynamic. When marketers talk about software products, they use terms such as features, components, products, suites, bundles, and so on. A group of benefits a customer gets, hence clubbed together as a talking point, are products.

Engineering

While all the other Microsoft Office products use Ctrl-F to search, MS Outlook uses F4 to search emails. It is an inconsistency highlighted by many MS Office users. When you have similar features in a product collection, people will expect them to behave alike. They would like to see tables in MS Word behave similarly to Excel spreadsheet cells. Considering these, Microsoft created engineering components that are across various parts of the products. Microsoft introduced binary components with a Component Object Model (COM) architecture. These could be developed by third parties and used across various products running on a Windows OS, thus achieving economies of scale in engineering endeavors. UI components and REST APIs used across products are similar initiatives engineering organizations use. Another aspect is to use open-source components. Since the company has not developed all the code for the open-source components, there are definite gains in reducing development costs. Some companies use common platforms or core technology components in many products. Again, the engineering gains are shared across products, thus achieving economies of scale. The claim is still hypothetical. While you can share components or platforms across the products, it's essential to ensure the components are capable and built to the needs of the downstream products. Additionally, the component releases must be aligned to the final product releases, avoid conflicting behaviors across the products, and remain resilient to the business agility required in developing products. All these add to engineering and lifecycle process management challenges. Today, most organizations use agile process management frameworks to realize such needs.

Infrastructure

We barely utilize 10% of the CPU, about 50% of the RAM, and hardly any network traffic when we type a document on MS Word. A laptop with an 8-core 1800 MHz processor is beefy hardware for such needs. Processing power, RAM, and even network connectivity costs have gone south as much as hardware prices. However, management of complex hardware, maintenance, and IT resource involvement adds to additional overheads. If we could utilize the hardware to its maximum potential, such additional expenses would be spread over the utilization. While typing this book, the CPU utilization is only 20-30 percent. If we run a test automation task alongside, the utilization will be 70 percent. However, that will not lead to a significant performance overhead for typing. Here are some simplistic anecdotal explanations of cloud computing:

In an Infrastructure as a Service (IaaS), we manage the laptop and keep usage ready; you can rent it on demand. You do not buy the hardware nor manage and maintain it.

In the case of a workload management system, we manage the resource availability in virtual or physical machines and place your Docker container. You get the uptime without bothering about the resource utilization concerns.

In a Platform as a Service (PaaS), you invoke the APIs and get the results. You do not bother about the underlying operating system, hardware, or system management tools.

In a Software as a Service (SaaS), you access a web page and type in your document without concern about the hardware or software involved.

There is progressively value added as you move from one stage to another. Hence, intuitively, SaaS will be more expensive than PaaS and IaaS. A SaaS vendor may depend on an underlying IaaS or PaaS vendor. For example, Netflix uses AWS extensively for its cloud computing needs.

The cloud reduces the need to stock up on hardware and software assets for peak load. You can rent as the demand changes. You start incurring revenue expenses and not capital expenses. For a startup, it is easier to justify revenue expenses than holding up funds in capital assets. However, a new customer acquisition adds marginal costs. You run the cloud infrastructure with slightly higher capacity to keep the new customer acquisition agile.

The second issue is with development usage or human interaction with hardware. When developers or authors work on hardware, they take breaks, are interrupted by corridor talks with colleagues, spend time planning or designing the content, and so on. During these periods, the resource is active but not in use. Hardware goes into a low power mode during these periods, but when you reserve a GPU or CPU in the cloud, you incur costs based on the total time you have reserved. In such cases, on-premise hardware may be cheaper for development needs. In the case of production use, you pay for the usage only. With significant revenues, the development overheads are not a concern in most cases, but for a small startup, any saving is welcome.

Working with IaaS, providing service configurations can be complex for most organizations. PaaS may be a quick way to start things. However, a developer may not be knowledgeable in all PaaS environments. They will only use the ones they have experience with. That may lead to vendor lock-in. Any vendor lock-in removes the flexibility of switching to a cheaper option identified later. Hence, SaaS model optimization for economies of scale can be complex, and improper planning may add to overruns.

Technology

The advent of Independent Software Vendors (ISV) in the ecosystem is a reality as the operating system vendors become flexible to enhance interoperability. When Apple decided to move its processors from PowerPC to Intel, they introduced universal binaries. These will run the executables depending on the underlying processor without extra effort from ISVs. Similarly, Google’s Flutter Framework can build client binaries and frontends for Windows, Mac OS, Linux, and the Web without significant code changes. Toolsets such as LLVMs can help write compilers and cross-compile code across various operating systems and processor types. Virtual Machine architectures in Java and Python-like environments have made software development OS-agnostic. The internet has made user experience operating systems agnostic, spreading the development costs across larger user populations. These are technological innovations that help the product mindset in software development.

The advent of AI has pushed the limits further. Companies such as Google, Microsoft, Meta, and Amazon provide platforms across products. For example, Google provides Gmail for consumers and enterprises. While developing a Large Language Model (LLM), the emails from Gmail can be used as inputs to the LLM. The prediction hints in a Gmail draft can utilize the LLM as much as Google Docs can to autocomplete sentences. Google Books and News can provide significant data to LLM development. In short, Google has developed a common platform where some of its products supply data. Some products can consume the data and provide insightful products and services. It is not confined to Google only. Every company developing a data platform is trying to reuse the platform across various products. Training an LLM may cost around 5 million USD.¹ A company investing in such a technology will like to reuse it across all its products in due course.

Figure 1.1: Cost of product spread across all the customers

We can say the software product is an innovative way to spread your costs across customers who view differentiated value from the similar solution you provide them. For our discussions in this book, this is the definition of a software product where product managers operate. The economies of scale in product development can only be achieved when the processes are managed with clockwork precision. Hence, there is a need for product management to work closely with the processes.

Product Management Process

Software product development and the penetration of PCs went hand in hand. The users were discovering the capability of the PCs and learning how to manage the new beast. The best product development companies employed their best technical minds to create innovative products, and then they were launched in the market. Even with massive launches, the market rejected many products. Marketing departments helped identify the target customer and set up price, promotion, and place, but they were not involved significantly in the product development. The void can be seen from the dotcom bubble bust. Every company offered a website-oriented free service to the users while advertisement revenue supported the service. Some companies succeeded using this approach, but large numbers could not survive. These failures made the industry wake up to the new reality and challenges. Some large companies looked at their software development processes and realized from development to launch, it had taken over seven to eight years, only to be rejected by the customers when they were in their 1.0 avatar. Some of the most successful products of our times had delayed launches, quality issues, and bad customer feedback in their first releases. The customer needs may evolve and change, and the product may not remain valid for the market. It led to organizations embracing the agile methodology for software lifecycle management. Agile methods require a product owner to represent the customer in product development; it ensures quick turnaround and market feedback, ensuring product relevance during the development cycle.

To support the product owners, the product managers had to interact with the market and acquire regular feedback. Some product managers played the role of product owners themselves. The e-commerce ventures were popular, targeting the business-to-consumer (B2C) markets. The product managers were data-driven. Since most of these applications ran on the web, they could insert probes to collect user experience statistics for analysis. In some B2C markets, the web-based systems are only digital conduits for other established businesses. For example, Walmart’s web storefront is just another channel for their physical stores. The product managers are responsible for the supply chains, existing systems, and processes, not just the web application. Business-to-business (B2B) customers were different. They acted as a conduit between the users and the product manager. The product managers had to engage as business analysts to collect customer feedback. The B2B world of product management involved negotiations, direct interactions, and feedback, while the B2C world relied on automated data collection. By the mid-first decade of the millennium, most client applications had moved to web browsers. Users start comparing websites or applications and expect a better user experience. Mobile devices and social networking websites took user experience to a much higher level of expectation. The scope of traditional UI designing started moving to providing elaborate user experiences, understanding user journeys, and providing the best solution. B2C software vendors focus on UI much more than B2B software vendors. Today, even B2B software vendors are focusing significantly on user experiences.

The need for developing competence in product management is identified in organizations. Organizations wanted the skill enhanced at all levels in product management. Industry consulting and training courses offered by Pragmatic Marketing², Blackblot Marketing³, and the International Software Product Management Association (ISPMA)⁴ came in handy to meet these needs. All these frameworks are detailed and define twenty or more functional processes for streamlining product management activities. Larger organizations find these frameworks handy as they establish a common understanding across product management organizations. The situation may not be as helpful for smaller organizations as they cannot formally train all the product managers, and there is a management cost associated with strict adherence to processes. ISPMA has introduced a relatively simplified process model defined for startups.

While we took a simple storytelling approach to describe the history of product management, researchers have conducted comprehensive academic studies around 2010 to define software products, the advent of software product management function, roles of software product managers, stakeholder management, the reference models or frameworks currently available in the field, and so on⁵. We suggest discerning readers review these articles for thorough insight into the product management process frameworks. A meta-ethnography study of the product management literature⁶ has identified the following core product management functions: product lifecycle management, product requirements engineering, release planning, roadmapping, and vision. It also identifies portfolio management, product analysis, strategic planning, product support, product development, and product launches as supporting activities. While the scope and definitions of product management were a set of discussion topics for the past decade, in the last couple of years, there has been a focus on finding the effectiveness of product management and the challenges the product managers face⁷. Although a small sample size may not be very significant to take a generalized call, the literature reviewed and the diversity of companies from where the samples were collected make the study apt for serious consideration. It may inspire researchers to work on related areas of study. They have identified seven areas where product management is facing the most challenges.

Requirement Engineering

Software Development Process

Team

Product Knowledge Management

Organizational - Strategic

Organizational - Operational

User Research

They identified 27 issues where three or more product managers raised concerns in an interview. The most frequently identified problem in the interviews was the team’s lack of motivation due to a lack of understanding of why they were doing this—resulting in no commitment to achieving goals. Similarly, the problem that got the highest severity and frequency was determining the true value of the product that the customer needs. Both of these raise significant concerns about how effective product management processes are currently. The researchers also worked with several product managers on potential strategies to overcome the challenges⁸.

The Framework

A search on LinkedIn showed 5.6 million product management personnel. When there is a surge in a particular discipline, there is pressure to quickly train them and make them aware of the product management function. When enhancing profits from the product is the focus of a product manager, engineering, customer engagement, or user experience are not the qualities for focus in a product management role. While larger organizations look at product management training programs to standardize the product management process, smaller organizations or startups question the very premise of product management processes. In startups, one or more of the founding members conceive the product. The person who has the best understanding of the product plays the role of a Chief Product Officer (CPO). However, the nuances of the product management process may not be there with such a person. The issue surfaces as a challenge in larger organizations as well. Product management is a vast domain. Not all people are well conversant with all aspects of product management. A person focused on sales support or customer evangelizing roles cannot appreciate the need for agile planning, product development, requirements management, and quality management. That is when there is a general tussle within and outside product management functions. Some organizations have junior members serve as product owners in the scrum process while product managers bring in market insights. Organizations that believe in such philosophies will not be successful when product dependency on technology is complex. A product dependent on cryptographic processing or a strong data-oriented focus may need a system architect-level person to drive the product owner role in scrum teams. A mere understanding of the customer’s requirements at a high level will do no justice to such a product. Many product managers have shown exemplary technical competence in their careers as well. Product managers pick up popular books such as Inspired⁹ and Empowered¹⁰ to learn nuances of product management and try to incorporate them into their day-to-day activities. Management programs and authors provide the perfect recipe for success when everything is followed in spirit. However, in many cases, the adherence is only ritualistic.

Figure 1.2: The Practice of Product Management Framework

Our approach towards product management is to establish a connection from product management functions to the management principles of sustained competitive advantage. The framework is also known as the five forces. We will deep-dive into the framework when we work on strategy. Michael Porter realized and showed with several case studies that an organization is continuously affected by five environmental factors, which he names five forces. They are:

Industry Rivalry

Bargaining Power of Buyers

Bargaining Power of Suppliers

New Entrants

Substitutes

Product managers orchestrate their functional skills to tame these forces. These activities are iterative. Product managers continually work with organizational strategy, cross-functional teams of people, and agile management to better these functions. We pick up each function in a chapter and explain how the function helps tame a particular force. For example, a good definition and execution of the Product Vision function can tame the industry rivals effectively. For bargaining buyers, the product management organization can move the buyer to a customer mindset, provide value-driven pricing, and improve the user experience for buying influencers. A well-negotiated contract reviewed in due course can help the bargaining supplier. Intellectual Properties can deter the advancement of innovative new entrants, while higher compliance requirements can affect a substitute taking over the market. An outline of the chapters is provided in the foreword. We call this framework The Practice of Product Management.

Conclusion

In our approach to simplify the product management processes for the experienced and the novice, we introduce The Practice of Product Management Framework. Reducing the overall processes to ten has made the framework lightweight for a practitioner. We add a causality element to the functional activities by associating the functions with the five forces. Lastly, we do not impose a firm process limitation. For example, we insist the organization fulfill the functions with their existing process and metrics if they have worked for them. A continual cycle of reevaluation provides a course correction and a new spin to the wheel. In the following chapters, we will work on these functions and utilize them in