Within the previous world, “pricing” was a factor you would level at.
A cellphone system value this a lot. A contact heart seat value that a lot. A migration value no matter your endurance may tolerate, multiplied by an hourly charge and a imprecise apology.
Now we’re pricing work carried out by techniques that don’t clock in, don’t ask for coaching, and don’t politely underperform till the third quarter. So we fall again to what procurement understands: seats, tokens, minutes, credit. Items which might be straightforward to bill, even once they’re onerous to consider in.
That’s the quiet shift Mark Vange, Founder and CTO of Autom8ly, is describing: consumers don’t buy “AI.” They buy the second the work is completed.
What Autom8ly Is Difficult About AI Pricing (and Why It Issues)
Autom8ly’s Mark Vange says token-based AI pricing breaks down in actual shopping for cycles as a result of it’s onerous to see, onerous to forecast, and troublesome to bundle for purchasers.
He argues that as AI strikes from “options” to accomplished duties, pricing will shift towards resolution-based models that clients can predict, approve, and funds—particularly in service supplier channels.
“The ultimate client pays for it by consequence… it doesn’t matter what number of tokens it takes.”
Why Autom8ly Thinks End result-Primarily based AI Pricing Beats Tokens for MSPs
Token pricing is rational in case you’re constructing an inner mannequin. It’s even rational in case you’re an AI staff making an attempt to measure prices exactly.
It turns into irrational the second you could promote.
As a result of the unit you’re promoting will not be “inference.” It’s not “GPU time.” It’s not “tokens.” The customer will not be strolling right into a funds assembly asking for approval to buy an unknown amount of invisible math.
They’re asking for approval to remove an issue.
Vange’s framing is blunt: clients purchase completed work, and the rest is vendor-centric accounting dressed up as transparency.
“If I are available in and measure it in tokens… it’s nothing to you.”
What’s occurring here’s a collision between two worlds:
The AI vendor world, the place value is granular, variable, and measured in micro-units.
The service supplier world, the place worth have to be packaged, repeatable, and defensible to a buyer who doesn’t need shock invoices.
That mismatch is why outcome-based pricing is enticing to MSPs and UC suppliers. It permits them to productize AI into one thing sellable. It turns “AI functionality” right into a line merchandise.
And crucially, it makes the dialog boring once more—which is strictly what procurement needs.
10DLC Onboarding as a Priced End result
Autom8ly’s instance is 10DLC marketing campaign registration within the US—one thing that’s necessary for A2P SMS, complicated for small companies, and labor-heavy for service suppliers.
The declare is that MSP onboarding for 10DLC can take 4–6 hours of hand-holding, whereas their AI method reduces that to a ~5-minute evaluate by the MSP, with ~95% first-time submission success.
As an alternative of charging “for AI,” Autom8ly prices per profitable marketing campaign utility. A set unit. A predictable SKU.
The deeper level isn’t simply that they automated a kind. It’s that they discovered a unit a buyer can conform to while not having to grasp how AI works.
“The value relies on this way getting submitted and handed.”
Associated Story on CX At the moment
Why Token-Primarily based AI Pricing Fails on the Shopping for Desk (Even When It’s “Honest”)
Token pricing has a seductive pitch: you pay for what you utilize. It’s measurable. It maps to value.
That’s additionally its drawback.
In most real-world shopping for committees, “truthful” will not be the highest requirement. Controllable is.
Finance groups need:
Forecasting accuracy
Spend governance
Contract readability
The flexibility to check distributors with out studying a brand new unit of physics
Token pricing struggles on all 4.
It’s additionally psychologically backwards. Tokens ask the client to care in regards to the vendor’s inner mechanics. However consumers don’t wish to sponsor your infrastructure selections. They wish to pay for their very own outcomes.
Even when distributors try to repair this with dashboards, the dashboards turn out to be a part of the tax: one other system somebody has to observe to forestall funds surprises.
End result pricing will not be excellent, however it has one huge benefit: it attaches spend to one thing legible. One thing a line-of-business proprietor can level at.
In case your AI can’t be pointed at, will probably be handled like a variable danger.
“With the ability to relate that operational value… to outcomes… makes it that a lot simpler so that you can undertake this.”
The UC Angle: Why End result Pricing Suits Service Suppliers Higher Than Consumption
UCaaS channels dwell and die by packaging.
They’re not simply promoting capabilities. They’re promoting deployment, change administration, compliance, adoption, and help. That’s why the “unit” issues extra for them than it’d for a hyperscaler.
The minute a UC supplier tries to resell token-based AI, three uncomfortable questions seem:
Who carries the overage danger?If utilization spikes, does the supplier eat the associated fee, or does the client get the invoice shock?
Who can clarify the invoice?Not “what induced it technically,” however who can clarify it in a method that doesn’t sound like a shrug.
Who owns the result?If the AI fails, do you refund tokens, credit, minutes, or one thing else?
End result-based pricing simplifies all three. It doesn’t remove complexity underneath the hood—however it prevents that complexity from leaking into the client relationship.
It additionally aligns with how UC suppliers already promote: bundles, per-site charges, per-location pricing, per-activation companies, onboarding packages, managed companies tiers.
End result pricing is just that mindset utilized to AI execution.
“Understanding the use case… consists of understanding the economics and the financial levers of that use case.”
What End result-Primarily based AI Pricing Actually Forces Distributors to Admit
End result pricing seems like a pricing tactic. It’s truly a product maturity take a look at.
As a result of when you cost for outcomes, you’re now not charging for effort. You’re charging for reliability. That forces uncomfortable self-discipline:
You want clearer definitions of “achieved.”
You want tighter guardrails on edge circumstances.
You want operational workflows that scale back exception dealing with.
You want a failure coverage that doesn’t turn out to be a margin crater.
Token pricing lets distributors cover behind variability. If the client complains, you’ll be able to say: utilization went up.
End result pricing removes that excuse. If the client complains, the implied query is: why didn’t the system full the work as anticipated?
That’s why consequence pricing, when it really works, is so compelling for channel companions. It turns AI right into a sellable product as a substitute of an experimental functionality with a variable meter operating within the background.
It additionally explains why the market retains circling again to the identical pressure: consumers need predictable spend, distributors have variable compute, and neither facet needs to carry the danger alone.
A current CX At the moment piece framed the parallel drawback involved facilities: pricing “seats” for brokers that don’t exist breaks down as autonomous brokers do extra work, forcing pricing groups to rethink the unit of worth. The unit is drifting away from individuals, and towards utilization and outcomes—as a result of that’s the place the work is transferring. Supply.
“It doesn’t matter what number of tokens it takes… the value relies on this way getting submitted and handed.”
The Believable Future Drift: “Decision Items” Develop into the New SKU
Right here’s the possible drift, if Vange is true, and if consumers preserve behaving like consumers.
First, the client stops asking “what number of tokens” and begins asking “what number of resolutions.” Not as a result of they’ve turn out to be extra subtle, however as a result of they’ve turn out to be extra drained.
Then the supplier packages AI the way in which they bundle the whole lot else: normal models, predictable tiers, tight boundaries.
A decision unit turns into a SKU:
One submitted and accepted 10DLC marketing campaign.
One compliant collections name accomplished with required disclosures.
One help case resolved with out escalation.
One onboarding workflow accomplished and verified.
This gained’t be marketed as “decision models,” at the least not at first. It’ll be marketed as managed automation, or AI-enabled companies, or “brokers.”
However the industrial logic would be the similar: consequence pricing quietly replaces compute pricing, as a result of compute pricing makes the client really feel like they’re paying in your uncertainty.
The darker drift isn’t dystopian; it’s administrative.
We’ll find yourself with contracts that look much less like software program licenses and extra like manufacturing agreements. Outlined outputs, acceptance standards, service credit, and countless negotiation over what counts as “success.”
And as soon as that occurs, the query gained’t be “how good is the mannequin?” It’ll be “how defensible is the bill?”
Not as a result of consumers dislike AI.
As a result of consumers dislike ambiguity with a fee schedule.
Sustain with UC At the moment
Be a part of the UC At the moment E-newsletter: https://www.uctoday.com/sign-up/
Be a part of the UC At the moment LinkedIn Neighborhood: https://www.linkedin.com/teams/9551264/
Should you’re seeing this pricing shift in shopping for cycles proper now, share what questions your CFO or procurement staff is asking—and what solutions they’ll truly settle for.
Observe the creator on LinkedIn
FAQs
1) What’s outcome-based pricing for AI?
End result-based pricing prices for a accomplished process or verified consequence moderately than for compute metrics like tokens, minutes, or GPU hours. The purpose is to align pricing with what consumers truly worth: completed work. It’s usually packaged as a hard and fast price per decision, submission, or accomplished workflow.
2) Why do consumers wrestle to approve token-based AI pricing?
Tokens are onerous to forecast and troublesome to map to enterprise worth in funds conversations. Even when token pricing is “truthful” from a technical standpoint, it could really feel uncontrollable to finance groups. That creates friction in procurement and slows adoption.
3) How does consequence pricing assist MSPs and UC suppliers promote AI?
Service suppliers want repeatable, packaged presents they will quote, ship, and help. End result pricing turns AI right into a predictable SKU that may be bundled into managed companies. It additionally reduces bill-shock danger and simplifies buyer explanations.
4) What’s an instance of outcome-based AI pricing in apply?
Autom8ly costs its 10DLC onboarding automation by the marketing campaign utility consequence—successfully a hard and fast price per efficiently submitted and accepted registration. That’s simpler for an MSP to resell than variable token consumption. The customer pays for completion, not mannequin effort.
5) Is outcome-based AI pricing at all times higher than usage-based pricing?
Not at all times. End result pricing requires clear definitions of “success,” robust reliability, and settlement on edge circumstances and exceptions. Utilization-based fashions might be less complicated to implement for distributors, however they usually create forecasting and governance points for consumers.
6) How far may outcome-based pricing realistically go if left unchecked?
If consumers standardize on consequence models, AI contracts could begin resembling operational service agreements with acceptance standards, credit, and audited definitions of “achieved.” That would scale back ambiguity, however it may additionally create a brand new layer of contract complexity round measurement and attribution. The “seat” could survive as a legacy wrapper, whereas actual worth shifts into priced outcomes.
