AI adoption for short-term letting: 8-step guide from the CEO behind 76% workforce adoption

Forge Holiday Group CEO Graham Donoghue closed out the Host Planet Roadshow in London with a fireside chat – then handed every operator in the room the playbook he uses to run AI across 26,000+ properties and 1,500 staff.

Why Graham wrote this guide – and why it matters now

During the Host Planet Roadshow in London, Forge Holiday Group CEO, Graham Donoghue, sat down for a fireside chat with Host Planet founder James Varley. The conversation centred on something most operators in the short-term rental industry are still wrestling with: how to move from talking about AI to actually putting it to work – across pricing, guest communication, owner management, and everything in between.

By the time Graham left the venue, he'd done something rare. He'd written it down. The result is a 15-page operator's guide built on the lessons Forge has learned running one of the UK's most advanced AI adoption programmes in travel and hospitality. We're hosting the full PDF here for the Host Planet community to download, free [scroll down to grab your copy].

The headline numbers behind the guide are hard to ignore. Forge Holiday Group – home to Sykes Holiday Cottages and Forest Holidays – manages over 26,000 UK properties. It has reached 76% AI adoption across more than 1,500 staff, one of the highest workforce adoption rates of any operator in the industry. That didn't happen by accident, and Graham's framework is a direct retelling of how it was built.

"AI won't replace great operators. But operators who understand AI will replace those who don't." – Graham Donoghue, Forge Holiday Group.

What's inside the 8-step framework

The guide is structured around eight steps, presented in the order Forge actually executed them. It's deliberately practical – designed to work whether you manage 10 properties or 10,000.

1. Build the data brain

Before any AI does useful work, your data has to be visible, integrated, and trusted. Graham's first instruction is to consolidate booking, owner, pricing, and customer data into a single source of truth – from a tidy Google Sheet for small operators up to Microsoft Fabric or Snowflake for larger groups – and to capture qualitative data (call transcripts, owner conversations) alongside the structured numbers.

2. Define your platform, access model, and architecture

Pick one enterprise AI platform (Copilot, OpenAI, Gemini or Anthropic), audit your existing PMS, CRM and channel manager for AI features you already own but haven't switched on, and standardise access tiers before you let anyone build anything.

3. Develop governance that enables, not restricts

Forge's principle: guardrails, not gates. Publish a short, visible AI policy. Maintain an AI register so every agent in production has a documented purpose, data source, and human owner. Brief leadership quarterly. People move faster when they know the edges.

4. Find your early adopters and establish an AI Lab

The best AI talent is almost certainly already inside your business – quietly experimenting with ChatGPT for guest replies. Find them, name them 'AI Guardians', and give them a cross-functional lab spanning Operations, Guest Experience, Owner Management, Finance, and Tech.

5. Build a use-case pipeline

Don't chase the most technically impressive use case. Chase the highest-frequency, lowest-judgement processes first – guest message drafts, listing optimisation, dynamic pricing, owner reporting, booking triage, predictive maintenance, review sentiment. The guide ranks each on business value, data readiness, and change complexity.

6. Optimise your talent and operationalise AI

Define clear AI roles – Agent Owner, Data Steward, Prompt Specialist, AI Guardian. Build tiered training (Foundational Awareness for everyone, Applied Practitioner for managers, Advanced Builder for data and systems people). Tie incentives to measurable outcomes. Graham's stated goal: a 30% competitive advantage from AI.

7. Publish an AI Value Dashboard

If AI's contribution isn't visible in margin and EBITDA, leadership won't keep funding it. Convert hours saved into pounds. Separate leading indicators (adoption rates, workflow automation percentage) from lagging ones (revenue uplift, owner NPS). Channel the savings back into more capability.

8. Begin building agentic AI systems

Once foundations are in place, move from AI tools that assist to AI agents that act. Forge's architecture connects large language models to enterprise data via the Model Context Protocol (MCP) – the control layer that lets agents plan and execute within governance boundaries. Start with one or two narrow agents. Document everything. Scale only what earns its place.

The four things that won't change

Graham opens the guide with a useful anchor. Whatever AI does next, four fundamentals don't move:

• People will still want to go on holidays.

• People will still want someone to manage their second home.

• Trust will always be a core element in decision-making.

• Technology will always evolve. AI is a step, not the destination.

Any AI strategy worth having is built on top of those truths, not in spite of them. It's a useful counterweight to the hype – and it's why this guide spends as much time on talent, governance, and human-in-the-loop principles as it does on agents and MCPs.

Why short-term letting operators should read this

The competitive stakes are clear. OTA algorithms are increasingly AI-native. AI-powered property managers are entering the market with dramatically lower overheads. Guest expectations – set by Amazon and Netflix, not by hospitality – demand instant response and personalisation. Goldman Sachs estimates AI could add 7% to global GDP over the next decade. AI agents are projected to make up over 60% of enterprise software by 2030.

For STR operators, two questions follow from that: what is your AI capability today, and how exposed are you to AI-native competitors who can scale faster and at lower cost? The eight-step framework is a structured way to answer both – from someone who's already done it at scale.

Download AI adoption for short-term letting

The full 15-page guide includes the framework in detail, the case studies behind it, the technology stack Forge runs, and the questions every operator should be putting on their leadership agenda.

Host Planet AI Adoption Guide.pdf

Download PDF • 50KB

Click the download link above to grab your free copy of AI Adoption – An Eight-Step Guide to Efficiency, Growth, and Future-Proofing Your Letting Business. Written by Graham Donoghue, CEO of Forge Holiday Group, exclusively for the Host Planet community after his fireside appearance at the Host Planet Roadshow, London.

Hear more from Graham Donoghue at the Host Planet Roadshow event in Chester

Interested in hearing more from Graham? He'll be joining us for the Host Planet Roadshow event in Chester – held at Sykes Holiday Cottages HQ – on 10 November. Tickets are available now – click here to get yours.

FAQs

Who is Graham Donoghue? Graham Donoghue is Group CEO of Forge Holiday Group, the UK parent company behind Sykes Holiday Cottages and Forest Holidays. Forge manages more than 26,000 UK properties and over 1,500 staff, and has reached 76% AI adoption across the workforce – one of the highest enterprise AI adoption rates in the travel and hospitality sector.

What is the 8-step AI adoption framework? It's a sequential playbook for short-term letting operators: (1) build a unified data brain, (2) select your AI platform, (3) establish governance, (4) find early adopters and set up an AI lab, (5) build a use-case pipeline, (6) optimise talent, (7) publish an AI value dashboard, and (8) begin building agentic AI systems. Each step has a defined goal, a recommended toolset, and a Forge case study.

Is the guide free? Yes. The 15-page PDF is free to download from Host Planet, written by Graham Donoghue and produced for the short-term rental community after his appearance at the Host Planet Roadshow.

Does this only apply to large property managers? No. The framework is explicitly designed for operators of any size, from 10 properties to 10,000. Smaller operators can start the data brain in Excel or Google Sheets. The principles – single source of truth, governance, early adopters, use-case pipeline – scale down as well as up.

What is MCP and why does it appear in the guide? MCP stands for Model Context Protocol – a control layer that connects large language models to enterprise data and systems. In the guide, it's described as the architecture that allows AI agents to plan and act on live business data within defined governance boundaries. Forge uses MCP to orchestrate its agentic AI layer.

What AI use cases should short-term rental operators start with? The guide recommends starting with high-frequency, low-judgement workflows: AI-drafted guest replies (reviewed before sending), listing optimisation, dynamic pricing tools like PriceLabs or Beyond, automated owner reporting, booking triage, predictive maintenance, and review sentiment analysis.

Is AI going to replace property managers? Graham's argument is that AI will not replace great operators – but operators who understand AI will replace those who don't. The guide is built on the principle of keeping humans in the loop, with AI extending judgement rather than replacing it.