Rolling out a rural network across Sub-Saharan Africa

In an increasingly competitive environment, middle market towercos are refining their rural offerings in order to work more constructively with operators. Among them is Africa Mobile Networks (AMN) whose zero-capex offering allows MNOs to economically extend rural coverage and comply with regulators’ coverage requirements. With many towers already built or in progress, their business model is galvanising an increasing number of deals in West and Central Africa as they stay on course to hit their target of 5,000 towers in the next five years. We caught up with their CEO, Michael Darcy, to find out how both AMN and the market have developed in the 18 months since they first shared their vision with TowerXchange.

TowerXchange: It’s been 18 months since we last spoke, can you recap AMN’s business model and tell us if you’ve made any changes to it since we last spoke?

We haven’t made any significant changes; we’re still executing to the same plan. We offer the operators zero capex and a revenue share on the base stations which we own and operate on their behalf.

TowerXchange: In our last interview you talked about 300 million Africans unserved by mobile networks and stated that 200 million of them will be reached within five years – how has this developed? Have these numbers changed much?

There has been some expansion of the existing networks of course but I don’t think we see the picture significantly different as there are still some 200 million people who are not served. Some can’t be economically served of course, but we say 100 million of the people not served could be served economically. Rural coverage just isn’t a high priority for the operators, most of them are focussed on 3/4G, transmission networks, fibre, capital investments and selling infrastructure to big towercos. Serving the rural populations isn’t very high on their agenda.

TowerXchange: What about regulatory pressure? Have actions taken by regulators on rural coverage benefitted you at all?

Yes, we do benefit, most licensees have coverage requirements and the vast majority of operators don’t meet their license conditions right now. In West Africa a raft of elections are coming up so governments are putting pressure on MNOs to meet licence conditions; it’s a vote winner. There have been threats to withdraw licences and impose fines on operators who don’t deliver on their rural connectivity responsibilities.

In fact, we even have examples of operators coming out of those meetings with regulators and calling us right away to rectify things.

TowerXchange: Please update us on AMN’s progress over the last 18 months.

When we spoke last we’d signed our first contract but not deployed any sites. We went on to build our first site in Benin which went into commercial service in January 2014. We’ve built a number of additional sites since then and are planning to construct around another 150 in Benin over next couple of years, so some sites are currently under construction, some in service, and a further 80 sites have been identified but takes six months to go through the process of building them. We’ll have around 150-170 sites by the time we finish but that will take another 18 months.

We also signed contracts in Cote D’Ivoire in 2014, and built our first base station there which went into commercial service in March 2015. We estimate to build around 150 towers there as well, rollout starts in earnest this summer. On top of that we’re aiming to be in three additional countries by the end of this year, so by the end of 2015 we aim for signed contracts and to be deploying in five countries in total. All of our clients in the three countries we’re planning to move into are tier one operators. Currently our focus is on West and Central Africa but our long term goal is to cover the whole of SSA.

TowerXchange: Your plan is to build around 150 towers in Benin and Cote D’Ivoire – would you say that’s an average number for AMN operations?

150 towers is a relatively small number which we would deploy in small countries. Of the three new countries we’ll launch into this year this year another is a similar size to our existing operations and another will require around 300 sites. The potential in Nigeria, for example, is ten times that. NCC and Universal Services estimate that Nigeria needs an additional 10,000 towers to serve the population. We have plans to build 2,000 sites there. Then a few countries are in between these kind of numbers. For us to be interested in going into a country we need to look at building at least 50-100 towers and we have turned down countries where there isn’t an opportunity to get to this scale. Our current plan is to build 5,000 sites in five years which is a drop in the ocean when you consider that Africa needs another 100,000 towers in the next few years.

TowerXchange: Have you managed to refine your site deployment methods and has it had an impact on capex for you?

We’re broadly delivering on what we expected to. It takes time to put a site in service – we commit to having the first site in service within three months which includes the core network integration that we need to do. Once that’s deployed we can add the rest as we need to. We then start the main roll out within six months. Towers and batteries et cetera need to be moved in by sea so that takes time. For the first tower, costs are considerably higher as we air freight everything in to get it up and running. We are aiming for ten sites a month per country, which isn’t much different from what we originally planned.

In terms of capex we are broadly where we expected to be. We’ve had to do some work to stay within those targets which has got more difficult as the US dollar gets stronger. But we need tight control on capex for our business model without compromising on quality. You can see our base stations are similar to other towers but lower capex. Also we need to focus on the upgradability – if a tower is too flimsy it’s hard to have an expansion plan for the future.  Often towers have to be replaced as they can’t stand the load any more. You need an upgrade path so you don’t have to rebuild from scratch.

TowerXchange: Has your revenue model been successful? Has it changed much?

The revenue share model is that we invest all the capex, it’s a full turnkey service and the operator does nothing. We install our equipment into their data centre and deliver all the revenue and do the full site evaluation and build, then do the full managed O&M as well. That’s our proposition: we deliver the traffic and we will take a share of the revenue we put into your network. Rural will always be very niche. It is meaningful to get another 4-5% of subs and revenues but 95% of revenue still comes from big towns and cities. In rural areas you might serve 10% of the population but ARPU is often lower. Incremental revenues and subscribers are guaranteed to be EBITDA positive. But of course signing contracts with big MNOs isn’t easy even when you have a good offering.

Our biggest challenge is that it’s a different business model to what people are used to. We haven’t had any problems getting to the right people, but the bigger the MNO the more complicated it is; some of these companies are huge with complicated decision making processes so it takes time. All of the companies have been very different to each other - some we’ve been talking to for 18 months with nothing signed yet but others have taken around six months even for tier one operators.

TowerXchange: Now that the majority of the tier one MNO relationships in Africa’s most dense urban areas have been sewn up by towercos, there does seem to be an opportunity for middle market towercos to work creatively with tier two and three MNOs for a new wave of tower activity. Is this something you’re seeing in the market?

When I talked to TowerXchange last I mentioned that our proposition was more compelling for the tier two to three and that’s a true statement. But like for the big towercos, the tier one operators have advantages in terms of financial stability. So we don’t focus only on tier one MNOs but they’re important to us. And every level is interested in our proposition. Our next couple of contracts will be with tier ones. They have the same constraints on capex as everyone else. Towercos have a lot to go at still and will focus on doing what they’re doing – rural isn’t high on their list of priorities either, which leaves opportunities for niche towercos like us to step into an area where there’s less interest.

TowerXchange: How has your fundraising gone since we last spoke?

We did indeed close the first funding round at beginning of 2014 with angel type investors. We’re currently funded by a VC fund – Harwell Capital - who have been funding us since H2 2014. By the summer we’re expecting to close a significant equity funding with a major private equity fund. We have two private equity funds who have indicated their interest through a term sheet to invest substantial amounts which we will leverage with bank debt and which is the main funding for infrastructure investment. We’ll close one by the summer which will fund new builds in all five countries.

For Benin and Cote D’Ivoire we have a long term loan agreement with SocGen for the debt pieces signed in last summer. Rural is becoming quite competitive and you can find half a dozen towercos who want to do the same thing. We’re very focussed and had a game plan from the beginning, we understand the technology well, and we have right access to funding and operators. All hugely important. We think it’s all about getting the pieces of the jigsaw in place.

TowerXchange: You mentioned you’d like to do an AIM listing when we last spoke, is this still on the cards and what do you feel you need to achieve to get there?

It is still very much our plan A which is once we have sufficient critical mass we’ll do a listing, probably AIM in London within two to three years of now.

TowerXchange: Please could you give us an overview of the passive and active equipment installed on a typical AMN site, and give us a comparison of the capex deployed relative to a conventional ground based tower, for which the passive infrastructure, installation and permitting alone can cost US$100-200,000 in SSA?

AMN base stations are typically installed with 2.75G functionality (voice, SMS, GPRS and EDGE data) with 2TRx capacity, and are expandable to 4TRx and 6TRx and beyond as traffic demand requires. We are also able to add 3G service when traffic supports the ROI. The antenna configuration can be omni-directional, but more typically involves 2 or 3 directional antennas. The passive equipment includes the tower – typically 30 metre or 45 metre – and fencing with a 10-metre x 10-metre footprint. The energy system is solar based with no reliance on grid or diesel power. The transmission network is based on VSAT. AMN is able to construct base stations with significantly less capex than a conventional implementation, and also with significantly less opex.