Taking a triangulated approach to monetising GenAI
• 6 minute read
By Simon Comfort, Pricing Specialist in Hg’s Value Creation Team
At our recent Growth Forum in New York, Simon comfort, Hg's Pricing Specialist highlights how SaaS businesses must rethink their pricing models and develop a robust monetization strategy to harness GenAI's potential and stay ahead in the competitive landscape.
As GenAI redefines value in the software world, SaaS companies face a pivotal choice: innovate their pricing models, as part of a well-executed monetisation strategy, or risk being left behind in the race for profitability.
There is little doubt that GenAI innovations represent an important growth opportunity for software firms, as well as a significant shift in the value that software can offer customers. Yet given the relative immaturity of the market, only around a third of companies who have developed such solutions are currently happy with their path to profitability. Careful consideration of packaging and price model, as well as price level, is vital as companies develop their monetisation strategies, with early customer engagement and feedback also necessary to ensure success.
One of the challenges that this technology presents is kinetic: the pace of innovation is such that the value delivered is rapidly evolving. As a result, customers may not always be clear on what that value is. Moreover, the costs associated with GenAI today warrant more significant consideration than with traditional cloud offerings (although we’re seeing costs declining sharply, at least for the time being). As this eats in to profit margins more than one would typically associate with a traditional software product – on average GenAI-related gross margins are ~10 percentage points lower due to model costs – it places even more importance on thinking about monetisation strategy early. Those who think this through at the product design phase, as opposed to making it an afterthought, and are willing to be nimble and adapt their approach, stand a good chance of getting ahead of the competition.
Pricing GenAI products remains relatively challenging given the factors above, but in many ways the fundamental approach is the same as for any other software. There are several key elements that should be considered:
Work out the ROI/value of the offering (theoretical maximum price);
Establish the cost base (theoretical minimum price);
Determine customer willingness to pay for the value delivered;
Compare to the cost of alternatives.
Triangulate cost, adoption and value
SaaS companies who embark on their GenAI journey are likely to find it necessary to iterate their approach over time, given the variables involved and rapid pace of change. Everybody is in the same boat in this respect. Software leadership teams should consider how they strike the right balance triangulating cost, adoption and value.
For example, some companies might decide early on to place more focus on adoption as they iterate a new feature and communicate its value to customers, taking on subsequent feedback to make changes. By moving fast through use of existing price models and competitive price points, this type of indirect monetisation strategy is less concerned about cost and value in the short-term.
As companies move towards a more medium to long-term view, cost and value-capture become a bigger consideration – although if costs continue to trend downwards this aspect will become less of a factor. Over time, customers will also gain a better sense of the value on offer. This will inevitably lead to an evolution in price models alongside price levels more commensurate with value.
Packaging…a key first step
Thinking through how a new product or feature fits into the company’s overall offering is a key first step, before determining the appropriate price model and price level. Some companies will introduce AI features to augment existing products while others might seek to roll out more transformational add-ons or even standalone, lead products.
Zoom’s AI Companion is available to customers on select paid plans at no additional cost. Such a strategy could simply be to keep up with the crowd, without being pivotal to the firm’s core product offering. Taking more of an indirect monetisation approach, whereby the main product is unlikely to change significantly over time, provides software companies a way to justify a general price increase or even just back up the overall value of their proposition.
Conversely, GitHub Copilot has proven to be more transformational in nature. Initially a bolt-on tool, GitHub Copilot has in effect become a lead product, used by more than one million developers. Last year, the company estimated that the productivity gains provided by GenAI developer tools could equate to an additional 15 million ‘effective developers’ by 2030, adding $1.5 trillion to global GDP. Clearly, it is tools like this that have the capacity to be directly monetised.
As the landscape matures, AI offerings are expected to become more integrated into software businesses' core offerings. This marks a significant evolution from the current trend of introducing AI solutions as optional add-ons with a price uplift, for example Microsoft Copilot (+50%), and GitHub Copilot (+90%).
Introducing an incremental AI feature to a core product, which adds a relatively small amount of value to certain users is a very different proposition to a potentially transformational standalone product.
A chance to disrupt pricing models?
One exciting aspect of the rise of AI-driven solutions is that it opens up the possibility for software companies to adopt a different price model and shift away from the status quo. Traditionally this has been thought about in terms of the number of users interfacing with software. But the traditional per-user pricing model is becoming less relevant: as software becomes ‘headless’ there may be no 'user' to monetise, and labour automation means users are likely a stagnant, if not declining, metric (albeit there is debate about how quickly we’ll see this happen). Instead, we expect a shift to more usage-based and output-based metrics, which better protect against margin risk – highlighted in part by examples such as Sam Altman’s recent X post, revealing that OpenAI is incurring losses on its $200/month ChatGPT Pro subscriptions – and can more effectively capture NRR growth with increased customer adoption.
Related to this, credit-based models, which scale with task complexity as different actions or use cases consume a different number of credits or tokens, are becoming more common, such as Clay or HubSpot’s Breeze Intelligence.
Companies such as Intercom and Zendesk are taking this a step further by using outcome-based pricing. This pricing model directly links to value and is becoming more feasible with the advent of agentic solutions that can complete entire workflows – allowing the specific impact from an AI agent to be tracked – and where the outcome is of a sufficient quality that the customer is willing to pay for it. For example, Intercom’s Fin AI agent is priced per successful AI resolution ($0.99), claiming to deflect 50% of a traditional customer support agent’s workload.
Hybrid adoption
Fully applying a usage- or outcome-based price model is unlikely to be achievable (or even desirable) for many companies, especially the latter given the difficulty in attributing and ascribing a value to outcomes. More realistically, we will start to see more hybrid adoption, where companies introduce some element of usage- or outcome-based pricing on top of their existing subscription model.
This approach provides the predictability valued by both customers and businesses/investors, while incorporating variable revenue elements to capture more value from increased consumption. For instance, a minimum commitment paid upfront plus overage charged in arrears.
There are additional considerations in implementing these types of models. Customer success will be increasingly important in driving upside through maximising outcomes and consumption. Sales compensation mechanics will need to evolve to account for consumption dynamics. Systems investments will also be required to track, report, and expose usage to billing and customers.
Whichever path software companies choose to tap into GenAI’s potential, careful thought is required at the outset. The scope for monetisation is substantial but the devil will be in the details.
This article is not intended to form the basis of any investment decision and is provided on a non-reliance basis. No representation, warranty or undertaking, express or implied, is or will be made or given in respect of the information contained in this article. This article contains forward-looking statements, projections, estimates and forecasts (together, “F-L Statements”). Not all underlying assumptions with respect to such F-L Statements have been included in this article. F-L Statements are not guarantees of future performance.