The Week’s Opener 
One hot question from us to kick off the week:
We want to hear the biggest hurdles to putting Outcome Based Pricing into action.
Some examples:
*Technical Feasibility: For some companies, will it even be possible to track outcomes accurately?
*Economics: Will the shift be too disruptive existing pricing models and revenue projections?
*Customer Inertia: Do customers even want to pay based on outcomes? Or do they favorthe predictability of existing models?
We want to hear from you - do any of these blockers resonate? Is there something else standing in the way? Smooth sailing?
Let us know how you’re thinking about this from your seat.
3 Likes
Based on what I’m seeing, I’m genuinely curious how many companies can technically implement outcome-based pricing. Given the influx of usage-based billing companies, it seems most of the SaaS landscape is stuck on legacy billing platforms, which could be a serious hurdle to innovating on pricing.
That said, Salesforce was able to launch Agentforce extremely quickly, which makes me optimistic that with enough momentum, others can pivot quickly too.
2 Likes
Attribution!
Ultimately, for outcome based pricing, you need to have a black and white measure of outcome. Most SaaS companies don’t have products that solve a problem in a binary fashion.
So, I think outcome based pricing is only a viable option for small portion of SaaS that can do proper outcome attribution.
5 Likes
We tried it, but for enterprises all this meant was not knowing how much budget they needed to allocate towards our product. Is it going to be 50k, 100k or 350k? Who knows.
Even having something like a credit-based outcome system wasn’t much better. So at least for enterprise the feeling is, predictability is still important.
Maybe once we’re more mature it’ll be easier.
1 Like
In my experience one has to jump three hurdles to implement outcome based pricing.
-
People can agree on the outcome and how it will be measured
-
The attribution as to what led to the outcome is clear
-
There is at least some level of predicability so that costs/revenues can be known in advance
If you can’t do all three of these things outcome based pricing is difficult and probably not appropriate.
If you can do these things the technical challenges can be met.
4 Likes
This is a great list @Steven_Forth - seems like 2-3 are the hardest part from my experience.
Yes, but it can even be hard to agree on outcomes for many applications. I have done a lot of work in learning and development and people have long and ferocious arguments about measuring outcomes.
I think rapid progress is being made on attribution. Here we have a lot to learn from healthcare where this is a formal discipline (part of HEOR or Health Economics and Outcomes Research). There is also a lot of progress being made on software tools in the Causal Machine Learning Community.
But for the next couple of years I think we will see outcome based pricing mostly in areas where these three hurdles are low or in places where it is already well established like Business Process Outsourcing.
I think predictability may be the easy. one. It is almost trivial to build a prediction model today. Most of the foundation models will do this and spit out the math and the code for you.
Absolutely right John. Attribution is a key issue. I’ve run into this during value modeling projects, i.e., what is really causing the uptick in outcomes? Your solution? Another solution? An externality?
Many products don’t have a clear causal effect.
There’s a saying “success has a hundred fathers and failure is an orphan.”
2 Likes
I think attribution is one thing and mentioned here as the main issue, but I don’t even think it’s desirable for most products.
Outcome-based pricing is only narrowly applicble. For some things it can work quite well, e.g. AI SDR, where a meeting is either booked or it isn’t and you can do attribution somewhat easily.
But many (most?) functions have no measurable, countable outcome. HR, brand marketing, engineering etc. don’t have a metric that correlates with value.
I think Goodhart’s Law is true here: As soon as a metric becomes a target, it ceases being a good metric.
You wouldn’t pay your engineers per PR or they’d stop working on building tricky features and start fixing minor UI bugs all day.
As AI becomes ever more capable to do things autonomously, we need to think about its compensation models.
And there’s a reason most human jobs aren’t commission-based, so why should AI doing the same jobs be paid a commission?
1 Like
We are right in the middle of a discussion whether we should get back to seat-based pricing. What is really challenging in our context is the fact that 99% of competitors price seat-based, so it is:
- difficult for customers to compare, and
- most importantly, estimate their needs and budget properly.
The perceived lack of predictability is really a show stopper for many. We are still doing more research to better understand why it is so hard for them to predict cost as in our mind if we have chosen the right outcome “unit”, companies would be able to predict that. Would be great to learn from others who have gone through a similar journey what helped you navigate the situation.
2 Likes
Uncertainty around predictability seems to be one of the biggest blockers from what I’m seeing.
Curious, Veselina, what is the unit you are currently charging for? Wondering if there’s a way to package it in a credit model to make costs more predictable.
On the first point, you need to move buyers from comparing price to comparing value. To do that you can use generative AI to generate value models for yourself and the competitive alternatives and then compare them to create a category value map and see where you are at an advantage (and where you are at a disadvantage).
Predictability is one of the three keys to outcome based pricing (to any pricing really). There are a few things you can do here.
- Use predictive analytics to build a prediction model for use. This is fairly easy to do these days. You need to do this for buyers and not expect them to do it for themselves.
- Design pricing so that there are floor and ceiling prices, keep these within a range that buyers are willing to accept.
- Have roll forward policies and make these flexible and favourable to buyers.
Hope that helps. Happy to discuss more.
2 Likes