Accelerate reliable, repeatable, traceable deployments across clouds and on-prem.

Guestline and Octopus Deploy

guestline.com

"We were able to scale quite well especially because of the fantastic support Octopus provides. It really did help us."

- Matthew Ford, Development Team Leader

Download Octopus Convert to PDF

Their Story

Guestline is a hotel operations platform that aims to provide an end-toend guest journey by using an intuitive resident property management ecosystem. They leverage the latest technologies to give guest time back to hoteliers by making manual processes integrated and seamless. Their products include a Property Management System and an Electronic Point of Sale (EPOS) software.

Guestline has over 2500 customers mostly based in the United Kingdom, however have expanded to other parts of Europe and APAC and have plans for further global expansion.

Challenge

One component of Guestline’s offering is their EPOS software, which is in almost half of the hotels they operate in. It is hosted across desktops and tablets and each customer could have a number of locations all requiring the software, often on around five machines per site.

Deployment processes were done manually and were a matter of copying and pasting code into each desktop or tablet in person. As a result, there were a number of inconsistencies in the installs as individual machines sometimes required a different version of the code despite being the same software.

This manual process made it difficult and labor intensive to update software as there was no way of making a new version available to every machine across all their customers without manual in person updates. It was estimated that rolling out a minor update to all their customer’s EPOS machines would take over 2800 man hours and around 350 days, costing an estimated £21,000, which is why it was done on a ‘need’ basis only.

Guestline saw the need to modernize this 20 year old process to improve customer experience and help their internal support teams that had to juggle an array of software versions. They wanted to automate this manual process so they could provide more frequent and consistent updates across their global customer base.

Solution

Guestline was already using Octopus Deploy on-premise for a few projects. Matthew Ford, Development Team Leader, made a business case for using Octopus Deploy to bring further automation to their process. Matthew had used Octopus in a previous role at a finance company, and was able to demonstrate the opportunity for further automation using Octopus, and how this could bring significant savings to Guestline.

Guestline was one of the earlier adopters of Octopus Cloud. They decided that Octopus Cloud met their needs for their EPOS system, and wanted to separate it out from their other uses of Octopus on prem so they had two separate environments. Cloud allowed the team to scale their support by managing hotels using multi-tenant deployments.

Having a production environment for every customer was not feasible in Guestline’s case with over 800 customers. So the team used tenants to automate updates across customers and set up each hotel as a tenant so they could send an update to one or multiple tenants at once. Using tenants also allows Guestline to deploy to any hotel on demand, as well as group hotels by region or customer group through filtering tag sets. This means that hotels owned by the same chain can be easily grouped together, and they can also use tags to identify customers who need bug fixes for certain software requirements.

Within the first few weeks of using Octopus, they had the first few customers set up. Over time they have been able to scale, bringing more of their customers over as they go. Now, over 650 of their customers are on Octopus, resulting in 4000 tentacles. There is around a quarter of their customer base left who are in the process of being moved onto Octopus deployments, which will bring them to around 5000 tentacles.

Guestline POS now has around two releases a month, and spreads this over a week or two. They deploy to their customers through an API that randomly selects and schedules around 100 customers to be upgraded at a time, with deployments taking around 10 minutes. This API has saved additional time the team would need to take in selecting customers for upgrade slots, further automating the process.

Results

Before Octopus, Guestline was not able to facilitate even small updates to the estate-wide EPOS software due to the time and cost involved in their manual processes, focusing more on targeted upgrades for specific problem installations. Now, using Octopus Cloud, they are able to facilitate hundreds of automated installations a day. Installations are now consistent across their customers, and are automatic without the need for anyone from the team being involved.

As a result, Guestline is saving 2000 person hours with each roll out, and has gone from no updates to one to two per month for its POS solution. This means customers are getting a world-class experience with their EPOS software, and the Guestline team no longer needs to be involved in manual rollouts.

With the process of keeping customers up to date being fully automated, it’s helped accelerate our plans to parts into the Cloud as we know we can get changes to customers quickly with very little effort.

Shout out to @OctopusDeploy for making their software so easy to work with. Just upgraded a 2 year out of date instance and migrated it to a new server and it worked with no effort beyond what their documentation said to do.

Twitter user Alex Dent Alex Dent
@DevOpsDent

We've been overhauling our internal infrastructure and back-end systems over the past month, including a move back to full @OctopusDeploy deployments; rediscovering how nice it is to have a platform-agnostic orchestrator that can deploy practically anything, anywhere ❤

Twitter user Nicholas Blumhardt Nicholas Blumhardt
@nblumhardt

Tools like @OctopusDeploy can be great in enabling culture change, we've been able to scale and improve our configuration story since we started using it.

Twitter user Niel Chalk Niel Chalk
@_neilch