Less than 2 hours ago, I was lucky enough to be able to tune in to the Vocus Communications Limited (ASX:VOC) results presentation while waiting for a flight in Los Angeles. A basic perusal of the press release would tell you that underlying NPAT was up by 53% to $13.6 million revenue was up 38% to $92.3 million.

CEO James Spenceley provided analysts with a little more detail when he presented the annual results this morning.

Importantly, on of the biggest changes is that Mr Spenceley now predicts that capex will remain roughly flat in the coming year. This is because the company sees "lots of opportunities," and as a result, Mr Spenceley believes that "artificially constraining capex would be a mistake."

Indeed, Mre Spenceley was "excited" to report that the company produced free cash flow of $9.2 million in 2014, even after the impact of growth capex. If the company stopped spending on growth, free cash flow would skyrocket to well over $25 million. I think Mr Spenceley said that "maintenance capex" was $3 million - $4 million per year, although I couldn't hear him clearly (the connection was a bit funny).

Vocus has four business segments, and an analysis of the company requires at least a glance at each of them:

Internet

Revenues were up a whopping 38% to $37.5 million, but the really good news was that the steadily reducing margins are beginning to stabilise. Interestingly, the company has begun to compete with its wholesale clients, albeit in only around 50 buildings. According to Mr Spenceley, "we’ll allow the wholesale client to compete and compete effectively."

Voice

Voice revenues declined by 16%. Vocus mainly offers voice as part of a product bundle and this is the least important segment.

Fibre and Ethernet

This division is the most important one because it provides the sustainable competitive advantage. Pleasingly, the revenue was up over 87% to $28.2 million. This division is to be bolstered by the acquisition of FX Networks in New Zealand. It is important understand that FX Networks has relatively low yield because the network has not been properly leveraged. A company like Vocus leverages a network by making it as dense as possible - think about the number of buildings connected in a given street, and the number of offices connected in a given building. Many value investors might otherwise think the price paid for FX was too high.

The company connected 407 new buildings to the fibre network, and is expanding outside of CBDs where economical to do so - for example they now have fibre running to Parramatta. Many of the 1048 buildings now connected have tenants who are not using Vocus fibre - and it is hoped that targeted marketing spend will result in many of these tenants joining the network at relatively low incremental cost to Vocus.

Data Centre

Data Centre revenue was up 19% to $18.6 million. Mr Spenceley noted that they expect further revenue growth from the Auckland and Melbourne data centres, and I was chuffed to hear that they are expanding one of the centres by removing the reception area. After all, data centres don't need a reception and that shows the company is serious about making the most of every dollar spent. Generally speaking, data centre purchases make sense because they allow Vocus to cross-sell other products. At least, that has been the experience in the past.

Finally,

The company is changing its marketing strategy. It aims to build its brand and "become a household name in corporate Australia." Ideally, this will reduce the cost of sales and improve the yield from existing assets.

The author owns shares in Vocus. The purpose of this blog is to document my thoughts on different companies in an easily accessible way and to make connections with likeminded investors. Subscribers to the Free Newsletter get sent research first, and have access to the Hidden Research.