The Process Behind
BUF Girl


BUF Girls is currently a functional, stable, profitable product serving hundreds of customers per month. Libby and Justin run it themselves and generate new content for courses with very little developer upkeep.

This is how we helped them do it.

The Client

Libby Babet, Founder of Bottoms Up! Fitness

Libby is the Founder of two of Sydney's best known fitness brands, AGOGA and Bottoms Up! Fitness. She is also the Co-Founder of Chief Bar, the Fitness Expert for Women's Fitness Magazine, the Fitness Expert for LifeStyleYOU and presenter for a number of health and fitness products and channels.

The Project

 The Bottoms Up! Logo.

The Bottoms Up! Logo.

Libby wanted to provide a commercially competitive online workout program which allows members of their bootcamp program to participate from anywhere in the world. She came to us with a plan to move toward a full product, a very solid understanding of her target market, and was easily able to communicate how she planned to stand out from the existing competitors in an already crowded online marketplace. This by itself is more than we normally receive from our clients, but Libby had taken it one step further.

Libby had already tested the basic concept prior to engaging with us, by running a very bare bones service of just manually sending content via email. Using this method she was able to gather a solid paying userbase, decide on the correct pricepoint, and receive key feedback which allowed her to fine tune her content. This gave us the confidence to move forward with the core idea quickly as validation had already occurred, allowing us to get straight onto refining wireframes and building a design after the initial meeting.

User Experience

We worked with Libby to create a simple user flow and feature set so we could launch as quickly as possible, our target was 3 weeks. We achieved this goal through UX focus sessions, wireframing, design, and then implementation. We felt this project really needed to be enjoyable to read and spend time in. So design and good user flow was important. Having programmers slap some code together with a sense of "making it pretty later" never results in that special feeling you get when something feels right. Sometimes the design is so simple it can be done on the back of a napkin. Sometimes it requires a process. In this case, we decided on loose wireframes...

Working this way allowed us to get a sense of the data model we needed, which we used to create an Entity Relationship Diagram and get to work on the data layer while we worked on implementing a design into a basic project. Within a few weeks we had a project running and ready to launch.



We believe choosing the tools which fit a project is far more important than picking tools which feel superior due to familiarity. Specialising in a few key areas is important, but ignoring new and emerging possibilities is an easy way to get stuck in a time warp with more stable,  productive, and generally better technology being released every day.

Git & Github - Code repository & issue tracking

Git is key to our development methodologies, we work in multiple branches and use a system similar to GitFlow to keep our repository ordered and understandable.

Github has become the industry standard for hosting Git, and we used a number of its core features  to our advantage in this project. 

Ruby on Rails - Language & Framework

We chose Ruby on Rails for this project due to the flexibility, ease of deployment, mature testing frameworks and the overwhelming amount of stable gems which fit our needs for this project. 


Amazon Web Services - Hosting & File Store

We chose EngineYard to manage our EC2 instances, which provided the server clusters, load balancing, error reporting and chef deploy process with minimal fuss. This cut down on a large amount of setup, configuration and maintenance which would be required for base EC2 instances, it also progressed quickly on a tight budget and allowed customisation which is difficult in some simpler but more expensive solutions (such as Heroku).

S3 was our go to file storage platform for this  project, good pricing and stability being the main reason for the choice. 

Postgres - Database

Postgres was our database of choice for this project, its stability, scalability and flexibility were proven in past projects. Full support for indexable arrays also fit an early need for object tagging, which made this an obvious choice over competitors.

Sass & Coffeescript - Asset generation

To quickly produce clean, reusable and small code we employed SASS (Syntactically Awesome Style Sheets) and Coffeescript to output our base assets. We find these enhanced languages which then compile down to raw CSS/javascript are much easier to initially write and maintain into the future. 

Bootstrap - Front end framework

We chose Bootstrap to get the MVP off the ground as quickly as possible and keep the initial budget to a minimum. Although some applications require a unique design to fulfil their MVP requirements, BUF was based around a few core pieces of functionality. Once the MVP had been proven, the default styling didn't stick around for long, but there are still some core elements which have only been modified slightly to fit in with the current design.