Jenée Hall — former CRO and pricing consultant who has run sales and pricing for SaaS companies across health tech, architecture, and agriculture — interviewed by Angelina on TwoSetAI (66 min). Her read: most technical founders are leaving money on the table because they treat pricing as a spreadsheet problem.

Pricing is an evolution, not a set-and-forget number

  • Starting from cost (“I know my token and API costs”) is fine as a floor, but pricing has to keep evolving with the product
  • Count ALL business costs, not just infrastructure — marketing, getting leads and eyeballs, support
  • A first price is a hypothesis to test against customers, then re-test; “we priced it this way and it’s been working” is not a strategy

Customers aren’t buying software — they’re hiring a digital employee

  • Ask what job the customer is hiring the agent to do, the way a manager staffs a team (“you wouldn’t put a PhD on a junior task”)
  • Mini-workshop case: a friend’s early-stage startup building an agent-orchestration layer claims ~5x cost reduction from self-evolving agents — but had no packaging, no pricing page, no landing page

The five pricing layers for AI agents

  • Access (per seat / workspace / org, monthly): classic SaaS and the least differentiated — pair it with a platform fee for onboarding, security, integrations, admin; that’s “right to use,” not where the value lives
  • Consumption (tokens, API calls, minutes, documents): useful internally, bad as a value metric — customers don’t understand tokens (“nobody buys kilowatt hours; they buy air conditioning”)
  • Agent capacity (5 active agents, 25 concurrent, 100 automations): mirrors AWS — cloud vendors don’t sell one computer, they sell compute capacity; cognitive capacity is the analog
  • Work completed (invoices reconciled, contracts reviewed, SDR campaigns done, tickets resolved): aligns price with “did the work get done?” — the outcome the customer can actually justify
  • Outcomes (+3% gross margin, 25 qualified meetings/month, % of savings): hardest to measure, highest value; revenue share / gain share / performance fees, with attribution kept clean
  • Hybrids work: a platform fee for predictable revenue, plus an outcome fee on top (“if we save you $50K/year, we charge less than 10% of that”)

Competitors as anchors, not answers

  • Research competitors’ price-to-value and position deliberately: price lower for market share, higher as a premium signal
  • Price skimming (Tesla: Model S first, Model 3 later) and penetration pricing (Netflix) are different plays for different moments
  • Anchor pricing done right is Steve Jobs’ iPad reveal: build the value case first — the price lands because of the lead-up, not the number; sales can’t just say “we’re cheaper than X”

Test price like a hypothesis

  • Talk to customer champions and ask willingness-to-pay the van Westendorp way — “would you pay $200? $250?” — to map the too-high/too-low range
  • Run experiments: offer the next 5 prospects $275; if they bite, that becomes the new baseline
  • Real example: client Monograph was charging ~$29–49 and testing found they could triple it
  • If you lower a price later, lower it for legacy customers too — protecting early high payers backfires on loyalty

The free-tier question

  • Free can build traction, but some buyers read free as low quality — and free users aren’t committed (“they expect it to tie their shoes”)
  • Charge something — even a $5–8 tier or a 15/30-day window — to filter out tire kickers; investors and boards will ask how many users are real customers
  • Freemium works when you make money another way (Google’s model, or Zoom cross-subsidizing small customers with larger ones)

Price should reinforce strategy

  • Pricing shapes customer behavior: remove friction for product-led growth, encourage multi-team adoption for enterprise expansion, premium positioning avoids the race to the bottom
  • Don’t copy OpenAI-style token pricing just because it’s familiar — usage pricing isn’t automatically right for your product
  • Design tier paths (e.g. add-ons that tip you into the next tier) for the behavior you actually want

“Buying anything is an emotional experience. Humans are still here because we are human and we have these emotions — no matter how much AI you have, we’re still the ones buying. You still have to appeal to the humans.”