How gen AI will reshape the software business

For all the impressive revelations and technical feats unleashed by the sudden emergence of generative AI (gen AI), one of the most astounding aspects has been the accelerated pace of its adoption, particularly by businesses. Consider that large global enterprises spent around $15 billion on gen AI solutions in 2023, representing about 2 percent of the global enterprise software market. To put that level of growth in perspective, it took four years for enterprise spending on the industry’s last major transformation—software-as-a-service (SaaS)—to reach that same market share milestone (Exhibit 1).

This unprecedented rise is just one indication of the massive disruption gen AI is poised to unleash on the enterprise software sector. Its impact could lead to a sizeable shift in user segments, value pools, and industry dynamics within and across software categories, presenting software leaders with vast opportunities and significant challenges. There are already indications of the extent of that disruption, such as the idea of agentic workflows replacing certain software applications The potential macro effects of gen AI are complex and interlinked, with some factors driving value creation and others fueling vendor switching, and in some cases, value erosion. This combination is likely to lead to a disruption—and in many cases a reimagining—of current software categories.

Of course, the full scope of long-term outcomes is difficult to predict, particularly given the unforeseeable evolution of the new technology and applications, and the myriad ways they could ultimately play out. But there is little doubt that a shift of some magnitude is on the horizon, so the early signs and implications of the gen AI realignment are well worth considering. This article explores these likely shifts and their broad potential impacts on the industry, based on a survey of software leaders and IT executives who make purchasing decisions, as well as our experience working with providers and buyers.

The coming value shake-up: Massive growth but greater disruption

Gen AI will drive significant growth in the software space. By 2027, spending on the technology could reach between $175 billion and $250 billion, contributing an additional two to six percentage points of growth for the sector. However, despite that sizeable boost, our research suggests the most lasting and disruptive impact of gen AI will be a wide-scale acceleration of vendor switching, on the order of five to ten percentage points. A range of factors will fuel this increase in turnover. New upstarts will likely be able to take advantage of the technology’s capabilities to erode established players’ advantages, with the decreased cost of data migration, integration development, and user training lowering the cost of transitioning off legacy systems. At the same time, software buyers are likely to give their legacy products stricter scrutiny as they race to stay current with the latest, most innovative gen AI solutions. In addition, total churn is likely to increase by one to three percentage points as the growing ease of software development spurs more enterprises to shift from buying to building their gen AI solutions, and certain software categories focused on data access and synthesis become commoditized (Exhibit 2).

Value creation: New use cases, easier development, and shifts in value pool and users

Almost all software categories are likely to have some impact from gen AI, although to varying degrees. New use cases and features driven by gen AI are likely to spur revenue growth. While most of the opportunity today centers around a few key capabilities—including a natural language interface to help analyze information, customized content creation, automated workflows, and enhanced unstructured data ingestion—this will likely evolve as the technology matures.

Even in these relatively early days, current enterprise adoption trends of gen AI use cases already provide a strong indication of the software use cases that will experience the most significant impacts. Perhaps it’s only natural that IT—the most traditionally technology-centric function—should account for the largest share of usage in this new category of software spending, at close to 40 percent. Use cases such as assisted code creation, IT helpdesk, and testing automation are already enjoying high adoption rates. Customer care is the other most active functional user at the moment, with use cases that include virtual chatbots and other customer care support solutions (such as script suggestions and transcript analysis) contributing to what we estimate will represent 15 percent of total spending by 2027. While other functions have yet to embrace gen AI adoption as fully, marketing and sales, along with certain parts of legal, auditing, and HR, should eventually make up a solid amount of the functional spending on the revolutionary technology.

Most of this spending on gen AI will fall into three main value pools, according to our survey (Exhibit 3).

Existing enterprise software solutions with embedded gen AI. Incorporating gen AI into established products could elevate the experience and productivity of traditional software, such as that used for customer relationship management (CRM), HR and human capital management, and productivity, which may or may not have any AI features. Typical use cases would be the introduction of natural language interfaces or agentic co-pilots for existing workflows.

Innovation and new products. Gen AI has enormous potential to spur the development of novel software applications. The product’s core features in this category are AI, with models created or finetuned for specific use cases or vertical workflows, especially when significant performance specialization is required (such as with drug discovery).

Enterprise internal solutions. The technological power of gen AI means that large enterprises will likely look to devote more resources to developing in-house solutions. These tools could be built for company-specific workflows, such as automating certain back-office processes or creating knowledge agents to help employees quickly access company data or information.

In addition to these three primary archetypes, additional spending will go toward tooling and infrastructure to support the application layer, such as foundation models and enabling technologies for AI development. These technologies could include AI governance tools, data storage and transport platforms, data visualization products, MLOps, or automation platforms.

Code development through gen AI is another significant source of value creation for software companies. Recent estimates indicate the technology can improve developer productivity by 35 to 45 percent, a spike that outperforms past advances in engineering productivity, leading to lower cost of code development. Gen AI can also speed up the processes of documenting code functionality for maintainability (which considers how easily code can be improved) by 50 percent and code refactoring by 20 to 30 percent. This dramatic reduction in the time and cost to develop innovative use cases and features, patch bugs, and continuously improve the offering doesn’t just drive margin improvements—it could have a significant revenue impact by ensuring software leaders are able to rapidly adapt to customer needs and deliver innovative solutions to them.

Software companies will also have to learn to adjust to a radically altered user base because of gen AI. This transformation has the potential to shift enterprise spending for some categories from traditional software users or labor pools (such as call center reps) to software applications (such as AI chatbots). At the same time, gen AI could expand the range of regular users of certain types of software.

In total, as much as 15 to 30 percent of knowledge work activities within each corporate function that adopts such novel solutions could be impacted by automation. Functions including sales and marketing, communications, design, finance, operations, and HR could see 20 to 25 percent of work activities affected by gen AI. Administrative, office support, and customer service functions, however, stand to be disproportionately impacted, given the sheer size of the workforces currently doing these roles, and the nature of the work gen AI can do. These roles revolve around certain types of software, including content creation, productivity, collaboration, CRM, and call center tools, and this is where gen AI’s value in bridging the barrier between human language and system language will be felt most.

Additionally, since gen AI can increase the accessibility of various types of expert software, the potential user base there could significantly expand (see sidebar, “How gen AI could expand the expert software user base”).

Vendor switching: Erosion of incumbency advantages and increased pace of replication

In the same way this revolutionary technology will likely upend traditional software value pools and user dynamics, our research suggests that gen AI may leave an equally complicated imprint on the larger forces that help determine success or failure in the business. Adopting natural language interfaces, for instance, could allow for faster onboarding to newer software, limiting avenues for maintaining competitive advantage within crucial software categories. Similarly, while software leaders are rightfully excited about the potential impact of gen AI on developer productivity, faster software development will mean competitors and upstarts can rapidly replicate offerings at a lower cost. Combined with the streamlined integration and lower switching costs enabled by gen AI, these trends have the potential to erode some of the built-in advantages industry incumbents have long enjoyed. We estimate the rate of vendor switching could increase significantly, potentially doubling, which in turn will likely drive greater competitive pressure on pricing.

To counter these shifts and the potential for increased competition, software players need to think ahead continually and innovate with new features. They could leverage proprietary data and insights as sources of differentiation. The ability to tap user experience as a distinctive element will still exist. Despite the availability of natural language interfaces, there will still be opportunities to differentiate with more finely tuned foundation models and develop more nuanced and specific user personas.

Value erosion: Increased internal builds and commodification

Similar to its impact of increased vendor switching, the potential of gen AI to improve the ease and cost efficiency of software development could cause enterprises to reallocate some software spending from buying to building their own products. Our survey indicates this impact would be relatively muted for the next three to four years, amounting to a two to four percentage-point shift in spending allocation. However, that would still be a substantial amount, at around $35 billion to $40 billion, and the share could grow significantly over the next decade as enterprise teams build more capabilities in-house and prove that they can build applications more expressly suited for their specific needs.

One notable factor that could accelerate the spending on internal builds is the potential increase of so-called citizen developers—employees who are not part of the core IT or tech group but who build applications and tools for consumption by themselves or team colleagues. Before the emergence of gen AI, the number of users seizing this opportunity did not accelerate as much as many industry experts predicted, mainly because low-code and no-code tools have had to overcome a learning and ease-of-use curve. Gen AI has the potential to unlock this type of software development in the coming years, with its ability to enable natural language-based application development. There are early promising signs, though enterprise-grade environments that could drive a step change are likely still a few years away.

Relative impact per software category

Based on our survey of software leaders and IT executives, we expect gen AI will have varying levels of impact across the sector. All software categories will likely feel some effects of value creation, switching, and value erosion, but the relative ramifications would be different. The consequence of all this change could be a disruption and reimagining of software categories as we think of them today. As exhibit 4 (below) indicates, we believe gen AI will cause every category to be re-imagined in some way. The following are examples of how this might play out:

• Certain software categories such as customer service could experience significant disruption driven by automation, reducing the number of users and changing customer engagement via gen AI assistants.
• Software categories for experts used in activities such as content creation or targeted to specialized teams could see significant disruption through the increased number of semi-expert users and the adoption of new use cases such as co-pilots that reimagine how users engage with the software. Our survey indicates greater potential for vendor switching in these categories.
• Software categories focused on enterprise automation will likely be transformed as they focus on shifting from legacy robotic process automation to becoming gen AI-powered hyper-automation platforms.
• Software categories such as CRM and enterprise resource management (ERM) that revolve around core enterprise workflows and contain sizeable amounts of enterprise or industry data could be considered ripe for value creation, given the increased potential to streamline and automate user workflows and leverage proprietary assets and insights.
• Software focused on ad hoc data querying, including business intelligence and visualization, will likely see disruption as gen AI supplants them and brings down the cost of tasks such as building queries, simple visualizations, or task automation.

Getting ahead of the disruption

In the face of such potentially significant levels of transformation, software leaders who want to stay ahead of the gen AI wave can start by asking some fundamental questions, including the following:

Are we moving fast enough? The pace of gen AI capability evolution and of adoption by both businesses and consumers suggests that a wait-and-see approach or attempting to develop a comprehensive strategy before diving in could be costly. First movers in this space could potentially capture much of the disruptive impact in the next 12 to 24 months. And those first moves might be relatively small from a product standpoint, such as incorporating a natural language chat interface, or rethinking a pricing model. An almost no-regret move, if not already underway, is to widely adopt and scale the usage of gen AI for software development.

Are we sufficiently reimagining our software category? Gen AI presents an opportunity for software CEOs to radically rethink what a software category means. It allows them to reexamine and address customer needs in a different way and potentially explore adjacencies that they could not before.

Have we strategically reallocated sufficient resources? Addressing this critical question depends to a large extent on a given player’s specific software category and the degree of disruption that is expected. But consider that if gen-AI-driven features and use cases seem on a path to contribute at least 10 to 20 percent of revenue within five years, an enterprise would probably want to devote a similar or higher share of R&D spending toward the technology.

Are we taking the proper steps forward? Capturing value from gen AI will require a concerted, cross-functional approach that includes regular input from across the organization, starting with the CEO. Forward-looking enterprises will want to consider altering how they price and package their products from quantity of users to amount of usage, as well as the possibility of new tiers for gen AI features. Another critical area will be figuring out novel ways to gather and leverage proprietary data, for example with product telemetry, and devising other kinds of product differentiators such as service offerings. Revamping product strategy and the road map for this new era will also be essential, with companies working to figure out how gen AI will impact their customers’ roles and workflows and using those insights to pinpoint the most compelling use cases.

The furious pace of gen AI enterprise adoption over the past year portends seismic change for the global software sector. Even though the revolutionary technology could accelerate industry growth, increase potential users, and fuel nearly $300 billion in new software spending within a few years, it also presents a range of new risks and challenges. Streamlined integration and lower switching costs that the technology will usher in could make it easier for new upstarts to grab business. At the same time, gen AI’s capabilities allow enterprise customers to build rather than buy more of their software. This disruption could translate to significant shifts in the industry’s user segments, value pools, and competitive dynamics. Software players that start thinking seriously about how to adapt to this fundamentally changed landscape will be much better positioned to thrive in a vastly new and different era that could leave some previously established leaders behind.