Insights

What growth-driven design (GDD) means for a hotel website

Growth-driven design swaps the big redesign every few years for a launch pad site and monthly, evidence-led improvement. Here is how it works for hotels.

Websites
Harry FielderCEO & Co-Founder
20 Oct 2026 · 7 min read

Most hotels buy a website the way they buy a refurbishment. A big project every few years, a launch, a sigh of relief, and then nobody touches it properly until it looks tired enough to justify the next one.

That is a print habit applied to a medium that can change every day, and it is why so many hotel websites are at their best on launch day.

Growth-driven design (GDD) is the alternative. It builds a website in three phases: a short strategy phase, a launch pad site that does the important things well and goes live quickly, and then continuous improvement in regular cycles, with every change based on what guests actually do on the site. The website is run as a product that gets better each month, not a project that ends at launch.

It is how our Website & Development practice builds and runs hotel websites. Here is how it works, what it asks of the agency relationship, and when it does not fit.

What growth-driven design is

HubSpot popularised growth-driven design as a reaction to the traditional redesign, with its months of planning, large upfront budget, late launch and a site shaped by the opinions in the room rather than by evidence. It has three phases.

  1. Strategy. Understand the guest, the business and the booking journey before anything is designed.
  2. Launch pad. Build and launch the smallest version of the site that does the important jobs properly.
  3. Continuous improvement. Run the site in cycles, each one choosing the changes most likely to move bookings, making them and measuring the result.

None of it depends on a particular platform.

Why the big redesign lets hotels down

The brief is written by committee. The project takes most of a year, by which time the offers, the photography and sometimes the general manager have changed. The site launches, the agency moves on, and the marketing team is left with a list of things that did not make the cut.

It is built on opinion

A redesign is decided before a single guest has used it. The room pages and the route into the booking engine are settled in meetings, and if the guess was wrong, it stays wrong for years.

The hotel keeps changing

A spa opens, the restaurant gets a new chef, a winter package needs its own page, a second property joins the group. A site designed around last year's business starts to bend as soon as it goes live, and every workaround makes the next one harder.

The technology keeps changing

When I wrote about the website lifecycle in 2018, the worry was new versions of PHP and browsers. The list is longer now: consent banners and Google's Consent Mode, Core Web Vitals, accessibility law, structured data, and AI assistants that read a hotel's website to answer guests' questions. A site that was current at launch drifts out of date unless someone keeps it current on purpose.

Left alone, a website only gets worse, and getting better takes regular, deliberate work.

Never settle for what you have. Nothing is ever a finished product. The launch pad is the starting point to evolve from.
Harry Fielder, CEO, Umi

The three phases in practice

Strategy starts with the booking journey

The tempting place to start is the homepage. I would start with the revenue path: how a guest gets from an advert, a search or an AI answer to a confirmed booking, and what is tracked along the way. That moves the brief from how the site looks to how the hotel makes money online.

Strategy also produces the first wish list, where every idea from the team is written down instead of argued over and waits for evidence.

The launch pad is a real website

The launch pad is the smallest site that does the important things well. Usually that means the homepage, rooms, the main offers, dining or spa if they earn revenue, and a booking journey that is fast, clear and properly tracked. It is a live website, not a mock-up, minus the nice-to-haves.

I would rather a hotel went live at 60 per cent than waited for 90 per cent and spent another season on the old site, because every month live is a month of real evidence.

We build launch pads on a content management system with a fixed library of blocks, so the hotel's team can add pages without breaking the design, and every new feature becomes a block any page can use. That is also why we avoid page builders on hotel sites.

Continuous improvement runs in cycles

After launch, the site runs in monthly cycles, and each one follows the same loop.

  1. Look at the evidence. Booking engine funnel data, analytics, on-site searches, heatmaps and what the reservations team hears on the phone.
  2. Form a hypothesis. Name a specific change and the result it should produce, such as fewer guests dropping out between choosing a room and paying.
  3. Prioritise. Score each idea on its potential impact and how easy it is to deliver, and take the strongest few into the cycle.
  4. Build and measure. Ship the change, measure it against the hypothesis, then keep, adjust or remove it.
  5. Feed it back. Pass what was learned into the next cycle and to whoever runs the ads and emails.

Nothing on the wish list is lost; it just has to earn its place. Priorities move with the seasons, so festive pages climb the list in late summer, not December.

A year after launch, a GDD site should look different from the one that went live, with a reason behind every change.

What the agency relationship looks like

In 2018 I asked whether hotels would buy a website more like a phone contract: a monthly fee for a site kept current and re-engineered as it goes, rather than left until it has to be thrown away. That is broadly how we work now.

  • Reactive support. Bug fixes, content help and anything broken, handled through tickets with response times set by severity.
  • Proactive roadmap. A set number of hours each month on the prioritised backlog, sized to how ambitious the hotel wants to be.
  • A regular rhythm. A weekly check on work in progress, a monthly review of performance and priorities, and a quarterly session with the wider team.

The first few months after launch tend to be more reactive, and after that most of the time should go on the roadmap. If nearly all of a hotel's monthly hours go on tickets, the platform or the brief is the problem.

It asks something of both sides. The hotel commits to a monthly cost instead of a lump sum every few years, and the agency has to show what each month achieved. In return, the agency has every reason to build a site that is easy to extend, because it will be doing the extending.

The Resident shows where that can lead. The relationship began with hosting their websites, then became a redesign, a full-funnel marketing programme and a direct-booking membership scheme, each step built on what the last one had taught.

Testing when traffic is limited

In theory, A/B testing sits at the centre of GDD. In practice, most individual hotel websites do not take enough bookings to prove that a new button colour made a difference, and Google Optimize, the free tool many hotels relied on, was shut down in September 2023.

So test changes big enough to show up, such as a shorter booking journey, a different room page or rates shown earlier, and measure them against bookings rather than clicks. Some decisions will rest on qualitative evidence, like session recordings of guests stuck in the same place, rather than statistical significance. Our guide to A/B testing a hotel website covers tools and sample sizes.

The aim is not to test everything. It is to stop making changes nobody can explain.

Testing is one part of a wider discipline, and our guide to hotel website conversion covers how to measure it properly.

When GDD is the wrong answer

Growth-driven design assumes there is a foundation worth improving. Sometimes there is not: the platform needs a developer for every edit, the booking journey vanishes onto a domain nobody can change, or the tracking is too broken to give any evidence.

Improving that site month by month only makes a weak foundation more expensive. That calls for a rebuild, and the signs a hotel website needs rebuilding are worth reading first. Even then, the rebuild should be the launch pad for the next cycle, not another five-year monument.


If you already have an agency, ask them one question. What did you change on our website last quarter, and what did it do for bookings?

If the answer is a list of closed tickets, you are paying for maintenance. If it is a list of changes with results, you are most of the way to growth-driven design.

Questions hotels ask

Growth-driven design is a way of building and running websites in three phases: strategy, a launch pad site that goes live quickly, and continuous improvement in regular cycles based on how visitors actually use the site. It replaces the big redesign every few years with smaller, evidence-led changes every month.

No. HubSpot popularised the method, but it is a way of working rather than a platform. Any content management system that lets you change pages quickly and safely, with proper tracking on the booking journey, can be run this way.

The cost is spread differently. The upfront build is smaller because the launch pad leaves out the nice-to-haves, and the rest is spent monthly on improvements chosen from evidence. It is an ongoing commitment rather than a one-off, and over several years it is meant to replace the next full redesign.

Conversion rate optimisation is the work of getting more visitors to book or enquire. Growth-driven design is the wider model for how the website is built and run, and conversion work is one of the main things that happens inside its improvement cycles, alongside content, SEO, speed and new features.

Yes, though the gains are limited by how much of the booking journey you can see and change. At a minimum the booking engine needs to pass booking data back to your analytics, so changes on the website can be measured against real reservations.

Free 30 Minute Consultation

Want a website that gets better every month?

Our Website & Development team builds hotel websites as launch pads, then runs them in monthly improvement cycles with every change measured against bookings.

Keep reading

More in Websites