Rethink, Retool, Reprice: An Agenda For Software In The AI Era

Recent headlines have framed Anthropic's latest release as the trigger for a sudden crisis in enterprise software, but that misses the fact that the share prices of many leading software businesses had already been weakening versus leading indices, reflecting concerns about valuation, growth durability, and competition. 

Sentiment has shifted from viewing AI as a straightforward growth tailwind to recognising that its impact on enterprise software will be nuanced; it is likely to create both winners and losers as it compresses margins in some areas, disrupts pricing models, and potentially erodes traditional moats, while opening new opportunities elsewhere.

A more grounded assessment of AI's influence on the enterprise software sector requires the examination of several underlying dynamics to understand how industry leaders can influence the trajectory in their favour.

First, incumbent software vendors still hold critical assets: customer trust, operational data, and an installed base of multi-year contracts. If they keep those customers satisfied, they have a meaningful runway to adapt – but time should not be mistaken for comfort.

Second, although no-code platforms and AI assistants are likely to enable customers to build their own systems, most organisations lack the time, focus, and strategic clarity to do so at scale. Vendors that help enterprise buyers maintain that focus while clearly demonstrating value are more likely to earn long-term loyalty.

Third, the magnitude of AI's disruption will vary widely across domains. In some categories – those with highly standardised, repetitive workflows – AI systems may outright cannibalise incumbent offerings. 

For instance, consider basic case reviews of suspected fraud involving low-value transactions, which large financial institutions have outsourced and offshored: these reviews are performed by third parties, based on specific protocols and well-templated processes supported by case management software, which could be replaced by agentic AI. 

However, in other areas, particularly where workflows are complex, customised, and judgment-heavy, AI is more likely to augment than replace existing tools. In these cases, enterprise software providers must make strategic decisions about how to enhance the value of their offerings.

Finally, economic reality will inevitably reassert itself. The current surge of AI investment cannot be sustained indefinitely without demonstrable returns. As the sector shifts more decisively towards monetisation, there may be an opportunity for established software providers with scale and durable customer relationships to become natural partners for AI challengers seeking to convert innovation into revenue.

The agenda for enterprise software incumbents

So, what should incumbents do? The answer lies in three parallel imperatives: integrate AI deeply, transform go-to-market approaches, and rethink pricing. While product integration is foundational, the most pressing commercial challenges lie in adapting how software is sold and monetised.

1. Incorporate AI across product offerings

Fundamentally, AI should become a pervasive component of the product offer, not a side feature. This must be absolutely aligned with customer priorities, be it efficiency (replacing or automating roles), effectiveness (improving throughput and decision quality), or both. The software sector has long been plagued by a “value gap” in which customers pay for features they rarely use. The disruption created by AI is an opportunity to shrink that gap, not widen it.

The “make or buy” decision also looms large. Some vendors may build their own proprietary agents; others may curate ecosystems of third-party agents or white-label agents from cash-strapped start-ups to gain speed.

2. Transform the go-to-market engine

Value selling is likely to be accelerated with AI, as functional buyers seek tangible benefits beyond core software functionality. A renewed product-led motion could emerge, with providers offering discounted or free AI add-ons to demonstrate impact before fully monetising the value they create. We're already seeing established software vendors move in this direction, such as HubSpot, which has launched a free AI Marketing Assistant. This will demand new skills, with a greater emphasis on demonstrating value through outcomes such as pilot-based proof, ROI calculators, and scenario modelling.

Product-led sales motions may change how customers buy. A 2025 Gartner survey found that 61% of B2B buyers prefer a representative-free buying experience, favouring digital self-service. While this may reduce reliance on traditional sales roles for lower-complexity products, enterprise sales will remain essential where integration risk, compliance, or multi-stakeholder alignment is required. The nature of sales is likely to change more than the need for sales itself.

Delivery models will also evolve. Successful AI adoption often requires workforce redesign and process re-engineering. This may lead to a shift in sales emphasis from new logos to customer success, ensuring that software is fully utilised and customer relationships are strengthened. It could also provide opportunities for partnerships, particularly with firms that bring operational and organisational expertise.

3. Adapt pricing models to new realities

The potential implications of AI on enterprise software offerings will inevitably affect their pricing. In addition, AI's impact on traditional SaaS economics must be considered, particularly rising variable costs tied to inference and cloud usage, as well as the efficiencies it enables in software development. These may call for new pricing approaches that align with evolving cost structures and margins.

Software pricing is evolving from price-per-seat to consumption-based and, more recently, to outcome-based models. The outcome-based model has a sound theoretical rationale for its application, as it aligns the value between the customer and vendor. However, for customers, it may be less predictable and more difficult to budget for. For vendors, too, it may be harder to implement. For instance, defining and measuring success is not always obvious and may present a real challenge for Customer Success teams and channel partners alike. These practical issues mean that hybrid models are likely to remain the preferred route for some time.

Determining the right price unit is equally important: should customers pay per agent, per AI capability, per API call, or per chunk of compute (token)? Similarly, entry pricing strategies will be critical to encourage trial while protecting proprietary data and customer relationships from outside AI agents.

Finally, all this feeds back into financial metrics. As variable usage replaces fixed subscriptions, metrics such as Annual Recurring Revenue (ARR) may lose some relevance. Investors and operators will need updated frameworks to measure and communicate growth, predictability, and underlying value creation in an increasingly usage-led world. We cover how valuation frameworks may develop in our recent 2026 Enterprise Software Technology Predictions Report.

Concluding remarks

AI will not destroy the enterprise software industry, but it is likely to redraw its economics and competitive boundaries. Incumbent software vendors still have meaningful advantages – customers, data, domain knowledge – but they can no longer rely on the old playbook. 

The imperative now is to align those strengths with the market forces reshaping the sector. The providers that systematically embed AI where it truly matters to users, reorient their go-to-market around demonstrable outcomes, and modernise pricing and metrics for a more usage-driven world will be best positioned to benefit as AI continues to evolve.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.