If you want alpha - find it at https://www.artisangrowthstrategies.com/

Joseph W.

Cofounder of PricingSaaS, a pricing intelligence platform. We help SaaS leaders make smarter pricing and packaging decisions through data, research, and community.

Rob Litterst

We tracked 1,800+ pricing and packaging changes across 500 SaaS companies in 2025.

That alone tells you pricing isn’t a once-every-few-years exercise anymore.
It’s something teams are adjusting constantly.

When you look across those changes, four shifts show up again and again (across AI-native startups and legacy SaaS.)

They say a lot about where SaaS pricing is heading next:


1️⃣ Credits became the default language for AI monetization

Credit models grew 126% YoY - with Figma, HubSpot, and Salesforce all moving in this direction.

They work because they sit between seats and outcomes: predictable enough for buyers, flexible enough for vendors to manage AI costs.

But the trade-off is obvious 👇

Credits add friction fast… (I think most of us have had at least one ‘wait, how many credits is that?’ moment. 😅)

Flexibility helps, but there’s a real pull back toward simplicity already.


2️⃣ AI is getting re-bundled into core plans

A lot of ‘AI add-ons’ are being absorbed into the base pricing model.

→ Notion, Slack, and Loom bundled AI and raised prices.
→ Airtable and Atlassian bundled AI with credit limits and held prices steady.

Either way, the message is the same: AI is moving from ‘extra’ to ‘expected.’


3️⃣ Seats are still here (AND customers are asking for them)

If I had a dime for every time I hear ‘Seat-based-pricing is done’… 😅

But buyers keep choosing it.

Seats feel familiar, predictable, and budgetable - especially compared to another new usage model.

What’s changing is what a ‘seat’ includes: more capability, more automation, more consumption, more flexibility.

In other words, seats are becoming a container for value, not just a login.


4️⃣ Buyers want options

AI doesn’t fit neatly into one model, so companies are offering options.

Salesforce now lets customers choose between seats, credits, PAYG, per-conversation pricing, and flat-fee enterprise agreements - sometimes within the same product.

Decagon goes further, letting customers choose which outcome they want to pay for.

Optionality is becoming a real advantage in the buying process.


I go much deeper into these datapoints, examples, and edge cases in a guest post this week on Kyle Poyar’s Growth Unhinged.

Should be helpful context if you’re thinking about where pricing is headed next in 2026.

Link’s in the comments 👇

Building the first stop for pricing and packaging.

Two companies can sell ‘AI credits’ and mean completely different things.

That’s because every credit model starts with two questions:

- Are your credits cost-based or value-based?
- How customer-friendly do you want them to be?

Answer those, and you almost always end up in one of four approaches 👇

1️⃣ AI credits as a product feature
(Value-based, vendor-friendly)

Credits are bundled into core plans, reset monthly, and abstracted away from models.

The goal is stickiness without runaway costs.

- HubSpot uses credits to create a growth lever beyond seats. 
- Adobe does something similar with Firefly, switching to pass-through pricing when customers use third-party models.


2️⃣ AI credits as a creative sandbox
(Value-based, customer-friendly)

Credits are pooled, flexible, and framed around what users can build.

Usage and experimentation matter more than tight margin control.

This shows up most in AI-native tools like Lovable or Replit, where AI IS the product.


3️⃣ AI credits as enterprise infrastructure
(Cost-based, vendor-friendly)

Credits track underlying costs closely and are tightly governed.

Caps, add-ons, and admin controls help manage usage volatility.

Cursor fits here: API-level pricing bundled into a seat. Simple to understand, but with no rollovers.


4️⃣ AI credits as a pass-through utility
(Cost-based, customer-friendly)

Credits map cleanly to vendors or actions.

Transparent, but unforgiving.

- PostHog is open about markup and allows partial rollovers.
- Clay goes all-in on credits, tightly tied to provider costs, with pooling and top-ups built in.


The general pattern I’m seeing?

Credit fatigue is already setting in.

As they continue to flood the market, customers will seek out flexibility that’s easy to understand.

And more companies are likely to drift toward value-based, customer-friendly models in the top-right quadrant.

Curious to hear how others are thinking about this?

AI is challenging a lot of the old assumptions around pricing and packaging - especially at the enterprise level:

👉 How do you price AI when outcomes aren’t guaranteed?
👉 What belongs in enterprise contracts and terms?
👉 Where does the AI product end and services begin?
👉 How should Forward Deployed Engineers be accounted for?
👉 How do you assess true willingness to pay?

In light of all this, PricingSaaS is hosting a small, office-hours session led by Ulrik this coming Friday (Jan 16 at 10 AM EST)

Ulrik works closely with teams navigating enterprise pricing decisions and will be there to discuss any open questions you might have - from structure and packaging to contracts and willingness to pay.

The session is capped at around 10 product, pricing, and GTM leaders, with a loose focus on where enterprise (and AI-driven) pricing is heading over the next year.

Expect plenty of discussion, minimal slides 😅

If you want in, just comment 'join' below and I'll send you the registration link.

See you there!

taplio

The Battle

Joseph W.

Rob Litterst

Followers

Posts

Engagements

Likes

Performance ScoreRob Litterst's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Joseph W.

Rob Litterst

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Rob Litterst's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreRob Litterst's LinkedIn Performance