My Reading List, 24.06.2014

Thought I’d share my [current] reading list. My blogging habits have not really changed, still intermittent.

Will keep updating this, it’s a fairly low blogging target.

 

 

 

ICT Policy Incongruence in Kenya

Kenya has made great strides in ICT in the last eight odd year.  We have moved from being what can only be considered 19th Century living (poor electricity supply, no internet connectivity, barely any telephone access etc), to being one of the regional ICT innovation hubs, with companies like Safaricom (Kenya’s telecommunication juggernaut, which is one of 4 companies in Kenya to have revenue of USD 1B + [KenolKobil, KPLC & Kenya Airways being the others], they are behind M-Pesa.), Ushahidi (who build crisis mapping software), the iHub, multiple incubators and more co-creation facilities. We have the capacity to innovate. We have a great developer base in Kenya. We have all it takes to be successful…

Software development is still not lucrative

Given the state of the market, this seems counterintuitive, but observe the numbers…. Which are our largest system integration outfits?

  1. Seven Seas Technologies –  One of the largest indigenous System Integrators. Revenue? As of 2010, 24 Million USD (with today’s exchange rate, 2.14 Billion KES, read more).
  2. The Copy Cat Limited -The oldest local System Integrator, they have revenues of over 50 Million USD (according to their website), with the current exchange rate, 4.45 Billion Kenyan shillings).
  3. Dimension Data East Africa  – Though it’s not Kenyan, strictly speaking, it is of Kenyan heritage and still partly owned by the Sameer Group. They went over the 1B KES revenue mark a few years ago.

Now, these are very well run companies, the developer community would do well to learn from them, however, my issue is with the government.

The government is the single largest customer for ICT products and services in the country (for a while the telco’s were ahead though, but given the recent USD 117.9M grant (with the current SDR Conversion rate) which translates to roughly 10.5 B KES, given the current USD/KES conversion rate). The governments policys should look internally for services and products to use before looking externally, especially in manufacture. This is because the government has a greater responsibility that transcends the need to increase uptake of ICT. The government has a duty to ensure that:

  • Our tax money is used to employ as many citizens as possible.
  • Government contracts are given to as many citizens as possible.
  • Fix our balance of trade.

Now, we have many multinationals setting up shop in Kenya (Oracle, Microsoft, EMC, Cisco etc etc). We have a large corporate market who should have no qualms buying imported product, their responsibility is to balance their books, source of products notwithstanding. However, this should not be the case with the government. The government is large enough to sway ICT trends.

Simple example. If the government standardizes say on FreeBSD for it’s server operations, all application vendors who want to deal with the government will have to port their application to FreeBSD or stand to lose business from the government. Simple enough. Given that the government is one of the largest employers, demand for UNIX administrators will necessarily increase (especially those with a bias towards FreeBSD). This will have a knock on effect because you will now increase the number of FreeBSD resources available in the market, making FreeBSD a viable server OS for corporations as there are enough skilled resources in the market to support it. The System Integrators will also ensure that they have adequate resources to maintain FreeBSD as they risk losing support and maintenance contracts if they are unable to competently manage the FreeBSD servers. Why is this important? Because we don’t have to spend a dime on FreeBSD. It’s free. We are paying the WorldBank loan with interest. It’s our tax money and the government has a duty to be as prudent as possible with the funds.

This plays out on a larger scale when you consider OS purchases, Information Management System (IMS) purchases etc. It’s a market that’s estimated by the Kenya ICT Board (MIS + Software and associated services) at roughly 200M USD. If this money was channeled to the local developer community, the  wealth creation will actually increase interest in the ICT space (it will be a viable career for the best and brightest) and will actually position Kenya to be a software *exporting* nation. I’m not against commercial software. I believe that companies like Oracle, Microsoft etc have many brilliant products that actually merit the government paying for them. All I’m saying is that paying for them should not be the first option. The first option should be the option that increases growth locally, even if it’s slightly inefficient. At the end of the day, whether the government buys an ERP from Oracle/Microsoft/SAP/OpenBravo, the final product is supported by Kenyan companies, with hefty fees paid to the mother companies, we might as well ensure that as much money goes into growing these organizations as possible. When we need to import software we might as well again ensure that the System Integrator who gets the final contract is a Kenyan company. This may be a bit tedious (involving breaking the tender into smaller manageable bits) but again, the government has a duty to grow Kenyan companies and keep as much of the citizen’s tax money internally…

So yes, we may be talking of improving the local dev ecosystem but the policy should actually actively aid the same, not pay lip service to the developer community and local entrepreneurs and leaving them in the dark when they have an actual business opportunity…

Hardware. Unexplored Potential

Kenya has a technology sector that has been accelerating over the last four odd years. We have had many startups move from oblivion into the mainstream and a lot of emphasis has been given to the new cutting edge sphere’s of technology. Some of these are:

  • Cloud Computing
  • Mobile Apps
  • Mobile Platforms
My problem? It’s great to have innovation around all these sectors, but the infrastructure required to run them will always be a source of money. Let me give a simple example. It’s not ‘cool’ to work on TFT Screens etc. However, an expertise in building TFT screens/CRT’s etc will eventually lead to the expertise required to build low power devices for devices such as the iPhone. Much as Samsung competes with Apple, Apple was one of their single largest customers last fiscal, contributing 5.7B USD to Samsung’s books (roughly 4%). Much as it’s good for Kenya to be leading in the mobile arena, it’s really unfortunate that we have not began innovating around the hardware side.
The ‘Kabambe’
The ‘Kabambe’ was a simple phone with only the most basic features that was developed by ZTE Technologies. It’ had a basic LCD screen with horrible visibility during the day. This was five years ago. Currently, ZTE are now OEM’s for Android. They have manufactured many high end android devices and they are one of the largest providers of hardware for Orange, both with regards to the devices and BTS infrastructure.
The ‘ZTE 225’ or Kabambe, as it was known in Kenya
How does Hardware Innovation Come in?
Huawei started off by running, then eventually building PBX switches targeted at SMEs. This eventually gave them enough legroom to launch their first digital switch (which at the time had the largest switching capacity in China). They continued expanding and eventually were able to start offering CDMA/GSM/UMTS, in roughly 1997.
This is similar to Nokia/Siemens/Ericsson/Motorola etc. Most of the companies that traditionally provided back infrastructure for PBX systems moved into the mobile space (it was only natural, fixed line communication was a dying business). Thing is, the skill-set required to launch and maintain a good GSM network works very closely with the skill-set required to make a good mobile phone (issues around network stability, power utilization etc). It was rare, at the time, to find a mobile phone manufacturer who did not deal in telecommunications infrastructure. The eventual entrants into the space (HP/Samsung/Acer/Palm etc) all had a previous background in electronic devices. We never really had a startup that was able to penetrate the smartphone market without either a telecommunications infrastructure background or a computing device background. As smartphones picked up, the manufacturers of said computing devices (HP/Samsung/Palm etc), grew at a higher rate as the devices took a a more end-user focused orientation which the companies had better experience in.
If you cycle back to ZTE & Huawei, they are now making amazing devices (both affordable and useable e.g.
The IDEOS, a USD 100 smartphone from Huawei
The ZTE Blade, a USD 150 Smartphone, available in Kenya
The Point?
In our Vision 2030 statement BPO is one of the pillars of the new economy. I believe (personally) that this is a waste of time primarily because one of the pillars of our economic growth is based on us maintaining low income rates for skilled labour (if we become expensive, we lose out), which is a conjecture, really.  I like the direction that has been taken with regards to software, but I believe more emphasis should be placed on hardware in Kenya. We will never be able to produce affordable phones before we are able to produce affordable computer monitors (the materials required are the same and the monitors reduce the cost of the raw materials in the economy). Why should we produce affordable phones? Because no matter how much software we develop, the customer will need to have a device to use the software/consume our media. Why can’t we be the ones manufacturing the devices? That will also give the local ecosystem an edge with regards to software development (the hardware devices will need to have some form of software be it from Microsoft/Google etc). If you think about it, Apple has expertise in both hardware & software and that’s what makes it a great company. All the great software companies today are acquiring a hardware skillset (Google/Microsoft/Oracle etc)… It’s time Kenya started thinking in the same direction…