12 Steps to Creating an Outcome-based Pricing Plan

Speaker A: What are the 12 steps to creating an outcome based pricing plan? Well, let's find out in today's edition of SAS Metrical. Welcome. My name is Ben Murray. So one of my recent blog posts, I'll have the link in the show notes because there's a lot of depth to this post but I'll cover the highlights of this the 12 steps to create an outcome based pricing plan. And this is based on analyzing, uh, lots of pricing pages out there, real live outcome based pricing pages and of course all the fine print, the terms and conditions behind that pricing page. Because we know this is complex not only to create but also to understand as the customer. How are we getting charged and when don't we get charged? What's that failure point where we don't get charged. So let's dive in. So number one, decide whether you actually have an outcome. This is a big one because we've talked about, well, CP based pricing is dead, subscriptions are dead. But actually is your software the customers you're serving, the workflow, is it built? Is it, could you have outcome based pricing? So ask whether your metric is tied to a completed customer result or just producing an activity so an issue resolved etc. So one we've got to determine do we have an outcome that we're achieving with our software or that we could achieve with our software? Step number two, define the outcome based pricing unit. The outcome unit is that atomic base billing event. It should be easy to count and tied to a workflow the customer already cares about. For example, Intercom's Fin AI agent charges $0.99 per outcome where an outcome is fin resolving an issue end to end or executing a configured procedure. So define that outcome based pricing unit. Step number three, define success criteria and define them twice for self serve and mid market products. Write success criteria like product line logic. There's a lot of logic in creating these in these plans. For example, help scout charges when AI answers and that's the product. M um, AI answers gives a full response and the customer does not ask for more help. It does not charge for greetings follow up questions without an answer or when the customer requests help. So define success criteria and then review that and validate that. Step number four, build in forgive failure forgiveness. This is a big one. Failure forgiveness is a conversion tool, just billing logic. So customers try outcome based pricing because they are insulated from execution risk. If the product fails, stalls, whatever it might be, the customer shouldn't pay. So ex, ex God, I can't even say the name of this company. An AI recruiting platform makes drop offs and no shows free. So for example, step number five, use a measurement window. Many outcomes are not knowable in that moment. We don't know a customer can look satisfied and reopen the issue two hours later. So for example, uh, for support cases like Zendesk, there's a 72 hour pattern where it counts an automated resolution after 72 hours of inactivity once its LLM confirms the issue was resolved. So you may have a measurement window just to make sure that that outcome was actually achieved. Number six, halfway through, decide whether to bill in dollars or credits. Some companies publish dollar prices, income or sorry, intercom $0.99 per outcome, help scout $0.75 per resolution. Others use credits to price across multiple outcome types. So are we going to charge, you know, dollars, cents or credits that equate to some dollar amount and then number seven, plan for the training leg. Outcome pricing assumes the product uh, produces uh, outcomes immediately, which may not be the case. AI products often cannot produce immediate outcomes. We could have training, we could, you know, going through, you know, you know, feeding all ingesting data to make the outcomes better. So Help Scout for example offers a free three month trial to train AI answers before charges begin. So we charge them right away or do we have a little training ramp up? Number eight, choose the commercial structure. Outcome pricing does not have to mean pure pay as you go. So you could have pure outcome based pricing like Intercom's fin 99 cents for outcome, you could have pay as you go resolution build in rel in arrears, you could have subscription plus outcome credits, you could have hybrid plus outcome and you could have enterprise custom outcome pricing. So a lot of different ways. It's not just rate times outcome volume. Number nine, anchor price to the type of value created. So separate value anchors into three categories. Labor savings, revenue generation and risk avoidance. Where are we creating that value? Number 10, almost there. Add spend controls before customers ask. In my previous podcast, the previous episode, talking about AI pricing trends and sentiment about those hard caps, if we reach usage limits, just like outcome pricing, we may only want to spend so much. But outcome pricing lowers adoption risk because customers only pay when it works. But it also creates a little budget anxiety. If it works really well, then we're going to be charging that customer more and more. This of course causes headaches for CFOs, for forecasting, for budgeting. Help Scouts let Help Scout, for example, lets admin set a monthly limit by the number of resolutions. And this is a big one. Disables AI answers when the limit is hit. So boy, talk about creating a pain point shifting instead of agentic a workflow, now the person's got to do it. Oh yeah, well let's pay some more instead of pushing it over to a human number 11. This is a big one, especially for the invoice for them to pay make every outcome audible. So outcome pricing creates a trust problem because the vendor defines, measures and invoices the outcome. If customers cannot inspect the billing events, they will not trust the model. So we need to it audible all the transactions, all those records to support that bill and then finally know when not to use outcome pricing. And I think this is the big thing just with usage, with consumption, not uh, I don't think everything is moving to outcome pricing. AgentIC AI pricing, it depends on the customer that you're serving. I just interviewed a founder today, was back on the podcast again after a couple years and they price uh, based on a percentage of value and this is in the construction industry. So know when not to use outcome pricing. So really quick, I have a lot more details on my blog post at the sass cfo.com, i'll put a link in the show notes if you'd like to go in deep on creating that framework, that structure to create outcome based pricing. Hope you enjoyed the episode today. Thanks.