AI Hospitality Alliance
Back to Insights

1 Jun 2026 | Terence Ronson

TCPG: Token Cost Per Guest

Pertlink's Terence Ronson proposes a new hotel KPI — Token Cost Per Guest — to measure generative AI spend per guest served. Drawing the parallel to how OTA commission moved from invisible deduction to board-level metric, he calls for USALI adoption before token costs sprawl unchecked across the P&L.

Tokens are the new commission

Every AI-powered interaction in a hotel — guest messaging, concierge, translation, upselling, review responses, RFP handling, revenue commentary, owner reporting — consumes tokens, and every token has a price. Ronson's argument: hospitality has seen this pattern before. OTA commission was once buried in distribution as "the price of incremental demand" until it grew into one of the most scrutinized costs on the P&L. Tokens will repeat that story, only faster.

Defining TCPG

TCPG = Total AI Token Cost ÷ Number of Guests Served

If a hotel spends US$120 on tokens in a month and serves 2,000 guests, TCPG is US$0.06 per guest. The number is small; its value is in what it reveals. Rising TCPG can signal deeper engagement, poor prompt design, or fragmented deployment. High TCPG is acceptable if it drives revenue, speed, loyalty, or labor productivity. TCPG is a value-efficiency metric, not a cost-cutting one — the goal is to optimize against measurable outcomes, not minimize.

Why it matters now

Without measurement, AI cost leaks across IT subscriptions, messaging platforms, CRM, revenue systems, departmental SaaS budgets, vendor bundles, and shadow tools. Ronson identifies five risks: cost leakage, poor ROI visibility, vendor opacity, weak governance, and guest-value misalignment. TCPG can be cut by property, department, use case, journey stage, or segment — and use-case TCPG is the most powerful view, because not all AI is equal.

A five-stage maturity ladder

Ronson lays out a path from Level 1 Invisible (AI used, cost untracked) through Token-aware, Departmental, Value-linked, to Level 5 USALI-informed — structured reporting alongside RevPAR, ADR, GOPPAR, and CPOR. He calls for HFTP, AHLA, and the Global Finance Committee to address AI token costs in USALI's 12th Revised Edition (adopted January 1, 2026), and proposes HITEC 2026 (June 15–18, San Antonio) as the venue to debate and benchmark the metric.

"Low TCPG is efficient. Productive TCPG is a strategy. Profitable TCPG is a transformation."

The real question Ronson leaves the industry with is not how many tokens were used — but whether they helped deliver better hospitality.

Read the full article on Hospitality Net →

Stay Updated

Enjoyed this insight?

Get more AI hospitality research and trends.

For Contributors

Submit article for consideration

Share an article, perspective, or industry analysis for possible inclusion in the Alliance Insights section.