In this episode of Hotel Moment, Karen Stephens, Chief Marketing Officer of Revinate, welcomes back Evan Crawford, Vice President of Marketing at Pyramid Global Hospitality, for his second appearance on the podcast. Evan returns with the same operational clarity that made his first appearance so valuable, but this time the conversation is grounded in the 2026 Revinate Hospitality Benchmark Report and the practical questions of what to do with the enormous volume of guest data flowing into hotel systems every day. His answer is consistent throughout: the opportunity is not in collecting more data, it is in connecting what already exists and using it with genuine intent.

● Guest feedback is now a generative search signal: Beyond its traditional role in operations, guest feedback is central to how LLMs assign authority to hotel content. Pyramid Global Hospitality treats reputation data as both an ops tool and a discoverability asset as the industry transitions from organic search to generative engine optimization.

● The incomplete guest profile problem: A customer who loved the hotel but had a terrible restaurant experience presents as a satisfied guest in the hotel data, unless the OpenTable review is connected to the same profile. Evan illustrates exactly why bridging these silos is not just a nice-to-have but a prerequisite for accurate personalization.

● Foundations before AI: Evan draws a direct parallel between the early days of SEO promises and today's AI vendor landscape, warning against "shiny objects" that cannot deliver without proper data infrastructure underneath. "If we're building a house without the proper foundation, it's going to crumble."

● Segmentation is the difference between value and noise: Reviewing the 2026 Hospitality Benchmark Report, Evan noticed that email volume and engagement are both high — but revenue through some channels is declining. His diagnosis: undifferentiated outreach is training guests to ignore messages. Each spray-and-pray send reduces the credibility of the next.

● Revinate emails still outperform ChatGPT on last-click revenue: Evan offers one of the most grounding data points in the conversation — at Pyramid Global Hospitality, traditional email and organic search still drive the majority of direct revenue. Getting ahead of AI is important, but abandoning proven strategies to chase emerging ones would be a strategic mistake.

● Content marketing was always the right answer: The shift toward LLM-optimized content is not a new discipline — it is a natural extension of content marketing done properly. "If we're just now scrambling to say, 'Oh my gosh, we need to develop content,' then we were doing it wrong for years before this."

● Writing for LLMs as an audience: Rather than keyword-stuffing, Evan describes the new discipline as writing rich content specifically designed to answer the prompts guests are likely to ask of ChatGPT, Perplexity, or Claude. Hotels need to think of the LLM as part of their audience, not just as a channel.

● Service recovery requires segmentation too: Sending a warm, promotional email to a guest who had a deeply negative experience is a failure of data activation. Evan's approach involves excluding detractors from standard marketing flows and instead routing them to personalized recovery outreach — sometimes a personal note from the GM — before re-engaging them commercially.

● Negative review responses are a sales tool: Evan reframes the audience for negative review responses entirely: in most cases, the response is not for the guest who left it — they may never return. It is for every future guest reading that thread. The right formula is to acknowledge the specific concern, take accountability, and close with a positive description of what the experience should have been.

Karen and Evan also discuss the upcoming NAVIGATE conference at the Wigwam in Phoenix, a Pyramid Global Hospitality property, and a keynote conversation with the Head of Google Applied Analytics on consumer behavior and the future of hotel discoverability.

If you enjoyed this episode, don’t forget to subscribe, rate, and review it on Apple Podcasts and Spotify. Instructions on how to do this are here.

Read the Hospitality Benchmark Report here: https://www.revinate.com/hospitality-benchmark-report/ 

[03:10] Guest feedback as a generative search signal — Evan opens with a reframe that positions reputation management as a forward-looking AI strategy, not just an operations tool. "We know that people are using the large language models more and more for discovery, and guest feedback is central to what the models are using to give authority to websites and to content. So we need to be prioritizing it to make sure that we're present and have a strong reputation when customers are searching a market for a product." For Pyramid Global Hospitality, the same data serving ops leaders is now serving discoverability in the LLM era.

[03:52] The incomplete guest profile — and how to fix it — Evan uses a concrete example to make the data silo problem tangible: a guest who loved the hotel but had a poor restaurant experience presents as a satisfied customer in the hotel system unless the OpenTable review is pulled into the same profile. "If we are not augmenting that guest profile to understand the complete customer experience, then we are just missing a key data point there." His argument: connecting platforms like golf reservations and OpenTable to the CRM is not a technical exercise; it is the difference between knowing your guest and thinking you know your guest.

[08:02] Shiny objects and the foundation that matters — Evan draws a sharp parallel between the early promises of SEO vendors and the AI vendor landscape of today. "There are a lot of shiny objects right now in the AI space telling you that AI is your savior. And what we're really trying to focus on is making sure that we've got the right foundation, whether it's our website data structured properly, or through our CRM and CDP. Because if we're building a house without the proper foundation, it's going to crumble."

[15:31] Tried and true still wins — One of the most grounding moments of the episode arrives when Evan checks the AI excitement against actual revenue data: "Obviously our primary source of revenue is not AI or the large language models. Even our Revinate emails drive more last-click revenue than ChatGPT does at this point." His point that AI should be ignored, but that preparing for AI and abandoning what works are not the same decision. "Yes, we want to prepare and get our house in order. But it's not just preparing for AI,  in many ways it's doing marketing the right way in general."

[17:36] Spray and pray is eroding your audience — Evan's read of the benchmark report leads to a pointed observation about how undifferentiated outreach is undermining the channels that still work. "We have so much data available to us, but that doesn't mean we should just spray and pray and just send mass marketing to people all the time. It should just be crafted in a thoughtful way to make sure that the customer actually wants to act on the message we're sending them." Each message sent without genuine segmentation, he argues, reduces the value of the next one.

Chapters:
00:00 - Intro

01:06 - Evan Crawford and Pyramid Global Hospitality's data approach 
03:10 - Guest feedback as an ops tool and generative search signal 
05:13 - Connecting data silos: OpenTable, golf, and the complete guest profile 
08:02 - AI foundations: avoiding shiny objects and building for the long term 
13:38 - Content strategy for the LLM era and what NAVIGATE will cover 
15:31 - Balancing AI preparation with proven revenue strategies 
17:36 - Segmentation, service recovery, and responding to negative reviews