Latest / MBA Training AI / Per-seat pricing: why agentic AI kills the seat

Transcript

Per-seat pricing: why agentic AI kills the seat · MBA Training AI

• 0:00
  Host: Welcome to the Leaders Insights Podcast.
• 0:08
  Host: Today's episode, how Agenic AI breaks the math behind your SaaS pricing model.
• 0:12
  Host: Every SaaS company on earth charges per seat.
• 0:15
  Host: You're about to tell me that's broken.
• 0:17
  Host: Convince me.
• 0:18
  Expert: It's not broken.
• 0:19
  Expert: It's dead.
• 0:20
  Expert: It just hasn't stopped moving yet.
• 0:22
  Expert: Per seat pricing assumes a human logs in and does work.
• 0:26
  Expert: Agenic AI, software that plans and executes multi-step tasks on its own, doesn't log in.
• 0:33
  Expert: One user now spins up 40 background agents that never touch your interface.
• 0:38
  Expert: You're billing for chairs in a room nobody sits in anymore.
• 0:41
  Host: Come on, vendors could just charge the agents a seat too.
• 0:44
  Host: Problem solved.
• 0:45
  Expert: They tried.
• 0:47
  Expert: It falls apart instantly.
• 0:48
  Expert: A human sales rep might send 50 emails a day.
• 0:52
  Expert: An agent sends 50,000.
• 0:54
  Expert: If you charge that agent one seat, you're giving away a thousand humans worth of work for the price of one.
• 1:00
  Expert: If you charge it like a thousand humans, the customer laughs and builds it in-house.
• 1:04
  Expert: The unit e-inter nido, the thing you actually meter, stopped being a person.
• 1:09
  Expert: And most pricing pages haven't noticed.
• 1:12
  Host: So what's the real unit now?
• 1:13
  Expert: Work done.
• 1:14
  Expert: Outcomes.
• 1:16
  Expert: The number of resolved tickets.
• 1:17
  Expert: Closed leads.
• 1:19
  Expert: Reconciled invoices.
• 1:21
  Expert: Intercom saw this early.
• 1:23
  Expert: They moved their support product to charge per resolution, roughly 99 cents, when their AI actually closes a case.
• 1:30
  Expert: Now, that's Intercom's own framing.
• 1:33
  Expert: And they sell support software.
• 1:34
  Expert: So take the 99 cents as marketing, not gospel.
• 1:38
  Expert: But the direction is right.
• 1:39
  Expert: You bill for the result, not the login.
• 1:42
  Host: Here's the uncomfortable part.
• 1:43
  Host: If I charge per outcome and my AI gets better, I close each outcome cheaper.
• 1:48
  Host: So I make less money as my product improves.
• 1:51
  Host: That's insane.
• 1:53
  Expert: That's the trap everyone walks into.
• 1:55
  Expert: And it's the smartest thing you asked.
• 1:57
  Expert: Yes, efficiency deflation is real.
• 2:00
  Expert: As the model gets cheaper, your per outcome margin can rise, but the price the market tolerates falls faster.
• 2:07
  Expert: OpenAI's own numbers show inference cost, what it costs to run the model once, dropping something like tenfold per year.
• 2:15
  Expert: That's their figure.
• 2:16
  Expert: They sell the tokens, so cross-check it.
• 2:19
  Expert: But even discounted, the trend means the thing you're metering keeps getting cheaper to produce and cheaper to buy.
• 2:26
  Expert: Volume has to do the heavy lifting.
• 2:28
  Host: So the answer is just sell more units?
• 2:31
  Expert: That's what every dying business says.
• 2:33
  Expert: No.
• 2:34
  Expert: The answer is to meter value the customer feels, not work you perform.
• 2:38
  Expert: There's a difference.
• 2:40
  Expert: O'Reilly's research on AI adoption makes the point well.
• 2:43
  Expert: The companies keeping pricing power aren't charging for tokens or tasks.
• 2:48
  Expert: They're charging for a business result the customer would have paid a human far more to get.
• 2:53
  Expert: If your agent recovers $100,000 in unpaid invoices, nobody argues about a $4,000 fee.
• 3:00
  Expert: If you charge them per invoice scanned, they'll haggle you to zero.
• 3:04
  Host: But that requires the vendor to prove they cause the outcome.
• 3:08
  Host: Attribution is a nightmare.
• 3:09
  Expert: It is, and that's the moat now.
• 3:12
  Expert: Whoever measures the outcome credibly owns the pricing conversation.
• 3:16
  Expert: Salesforce is betting the whole company on this with Agentforce,
• 3:19
  Expert: charging around $2 per conversation their agents handle.
• 3:23
  Expert: Their number, they obviously want agents everywhere, so treat it skeptically.
• 3:28
  Expert: But notice what they did.
• 3:30
  Expert: They picked a countable event a customer can see happening.
• 3:33
  Expert: The vendors who survive aren't the ones with the best model.
• 3:37
  Expert: They're the ones who can stand behind a bill and say, here's what we did, line by line.
• 3:42
  Host: And the ones who can't?
• 3:43
  Expert: They keep selling seats to a workforce that no longer exists.
• 3:47
  Expert: It's like charging admission to a stadium after everyone started watching from home,
• 3:51
  Expert: technically a business briefly.
• 3:53
  Host: All right, someone's running a SaaS company today per seat,
• 3:57
  Host: watching agents eat their model.
• 3:59
  Host: What do they do Monday morning?
• 4:00
  Expert: Pull your usage data and find the one moment in your product
• 4:04
  Expert: where a customer clearly gets value.
• 4:06
  Expert: A ticket resolved.
• 4:07
  Expert: A report shipped.
• 4:09
  Expert: Money moved.
• 4:10
  Expert: Count how often it happens per account, not per user.
• 4:14
  Expert: If that number is climbing while your seat count is flat,
• 4:17
  Expert: your pricing is already lying to you.
• 4:19
  Expert: Reprice against that event before a competitor does it for you.
• 4:23
  Expert: The unit was never the person.
• 4:24
  Expert: It was the outcome.
• 4:26
  Expert: It always was.
• 4:27
  Expert: We just had humans in the way hiding it.
• 4:30
  Host: Sources for today's episode, the decoder, TechCrunch AI, OpenAI,
• 4:35
  Host: vendor, AI Lab, Kated Nuggets, O'Reilly Radar.
• 4:39
  Host: That's your five.
• 4:40
  Host: The full AI track, structured, no fluff, is at MBA-training.com.