Monthly Archives: June 2014

On Its Way Out, Commercialisation Australia Extends Our Grant

Back in December 2012 we wrote about our success with the Commercialisation Australia (CA) program, after we were awarded a grant of $385,000 to help us commercialise our technology. 

CA has  now approved a significant extension, for a total of $1.2m, to our original grant. The funds will be used to bolster our development team, underpin the continued growth of our content, and generally assist the process of making Rome2rio the world’s leading multi-modal travel search engine.

The timing of this grant could not have been more finely tuned: we made our application just after a Federal Election, with the new Government hinting that they would review the entire CA grant scheme, and possibly eliminate it altogether.

Our application went to the CA Board  in February, at a time when CA themselves did not know their fate. We were in the dark not only about the Board’s decision, but also about whether it would be funded even if it had been approved.

We were not to find out our fate until some  four months later, and only after the new Government had announced that CA would indeed be eliminated and its functions assumed by the newly formed Entrepreneurs’ Infrastructure Programme (EIP). That news seemed to suggest that we would not be receiving our grant extension, but shortly afterwards we learned that all grant applications that had been approved at the February meeting would be fully funded. Talk about just scraping in!

As “outsiders” to the process of Government it’s tempting to look at changes like this as being driven by political rather than practical motives. With details of the EIP still emerging we can’t make any observations or comparisons to the now defunct CA scheme, other than to say we viewed CA as a well-managed, relevant and practical programme that clearly assisted many Australian companies, including Rome2rio, to grow and compete on a global scale. If EIP can improve on that, we’ll be the first to applaud.

Advertisements

Rome2rio retreats north to Townsville

Last year the Rome2rio team decided to escape Melbourne’s drab winter for a week by setting up shop at Palm Cove (near Cairns in Australia’s far north). That company retreat resulted in the development and launch of Rome2rio’s hotel feature, which has been a big success and has since become an integral part of the site.

Rome2rio's integrated hotel view was developed at the 2013 company retreat

Rome2rio’s integrated hotel view was developed at the 2013 company retreat

This year we decided to do it again by spending two weeks in Townsville. The Australis Mariners North Apartments provided an ideal base for work and Townsville offered plenty for the team and their families to do in their spare time, such as trips to nearby Magnetic Island and walks up Castle Hill.

A day at the office in Townsville

A day at the office in Townsville

We took the opportunity to focus on ideas, features and analysis that we wouldn’t normally have time to work on. The most significant project was the development of a new sights & activities component for Rome2rio. By the end of the retreat the team had developed a compelling prototype, which we plan to ship soon.

An early prototype of our attractions layer

An early prototype of our attractions layer

The second major project was the addition of transport images to Rome2rio. As our image gallery develops, Rome2rio will provide informative pictures of trains, buses, ferries, bus stops, stations, scenic views, ticket booths and much more. We also expect to launch this feature soon.

Showing transport images for certain transit lines.

An early prototype that shows transport images

We also took the opportunity to conduct a relevance analysis. Similar to an approach we used at Bing, the team divided into two groups and spent the day studying Rome2rio results for around 100 user queries that were drawn from our user logs. The search results for each query, such as Zurich to Bologna, was graded on a four-point scale of Perfect, Good, Fair or Bad. Relevance problems were categorized into buckets such as geocoder problems, missing transport data, inaccurate pricing or routing engine problems. The exercise is invaluable for prioritizing improvements to Rome2rio and is something we expect to do regularly.

Other projects included the development of a new A/B testing framework, experiments around new UI, conversion and SEO treatments, and new approaches to visualizing our query logs.

 Michael Cameron