In this week's post, we address a question that we expect is top-of-mind for every Cloud and SaaS executive-determining how to charge for newly developed AI features.
We are planning to announce a new service offering to address this in early October. In preparation for this, we have been surveying and interviewing our C-suite executive panel across industries on what they want to see from their Cloud and SaaS vendors. Here are some early voice-of-customer insights that may help with the pricing decisions:
- Technology decision makers expect their software vendors to provide certain AI enhancements to their existing products for free. These fall into the category of ongoing product enhancements that they feel entitled to as part of their subscription agreements. Let's call these AI infusions. AI infusions use AI to facilitate performing simple tasks such as creating summaries, creating basic commands and prompts to initiate tasks, etc. These tasks are still primarily performed by the user, with AI as a facilitator. Using the JTBD (jobs to be done) framework popularized by Clay Christensen, author of "The Innovator's Dilemma", these tasks can easily be mapped out using the following rule that our panelists appeared to reach consensus on. If AI is merely enabling the user to save their time or cognitive load in performing the task, then it is merely an AI infusion and is expected at no charge as part of the product update cycle. A rule of thumb is that if the task's cognitive load is 70% borne by the user and 30% by AI, then no upcharge is expected for it. Likewise, if the time savings from the AI infusion are approximately 30%, no charge is expected, although some on our executive panel placed this threshold at 40%.
- The second category is AI add-ons. Unlike the AI infusions described above, these add-ons bring a whole new capability to the product suite that the pre-AI product simply not capable of doing. Examples include a marketing automation suite that converted one form of content into another, such as converting this post into a landing page or social media ad. Using the percentage thresholds above, if the AI feature performed 70% or more of the cognitive load and/or time for the task, then it is a monetizable AI add-on.
- When asked for their expectations on what additional charge for new AI add-ons would be acceptable, 30% emerged as the average. Remember this number is what they would prefer to pay and not what they do pay or have paid. On the latter question, many expressed dismay at being charged up to 50% for these AI add-ons, along with dissatisfaction with user experiences and return on the investments. It appears that software vendors have rushed to push out still half-baked AI add-ons to their customers. To this end, decision makers are also asking for better guidance on roll-out strategies that pinpoint which users in the total user pool need the add-ons to avoid overbuying. Effective end-user training is another gap they identified to derive better ROI that they could demonstrate to their leadership.
So here again with AI, our software and cloud clients are faced with age-old price optimization question: How much to charge so that it is attractive enough for customers to adopt without leaving money on the table?
The visual below which I shared in an earlier post is from a recent client use case (identity redacted). Note the dramatic flatlining of the curve in a couple of places within what customers considered to be acceptable pricing ranges. In these parts of the curve, the company can charge more at the end of the flatline without losing “share-of-preference” which is a proxy for expected market share in conjoint/discrete choice modeling.
Alan Nazarelli is Founder & CEO of Silicon Valley Research Group. Based in San Jose, CA with offices in Seattle and New York, the company works with the world’s most innovative brands to provide timely and actionable market intelligence and strategic guidance to enable them to make well-informed decisions to positively impact revenues and profits and to achieve their growth targets. Connect with Al on Linked in