[This post has been edited from its original form. Its current form is simply a set of questions based on 7 years of UNICEF’s Innovation Unit experience in emerging market technology work]
Based off our principles of innovation (www.unicefstories.org/principles) The UNICEF Innovation Unit has created a series of 10 questions that can be explored by anyone thinking of doing a mass distribution of hardware.
Ten Key Questions to Answer Before Distributing Handsets or Other Hardware:
- Do you really need to distribute these devices? Why? Have you worked with local technologists to do some quick prototypes of how these devices will be used? Have you tested them locally?
- Who are they going to? Have you worked with local designers to do some quick, iterative user testing with your end users to make sure this is the right solution?
- How many of those people (x/y) already have a smart-phone? How about a dumb phone?
- How will you account for loss? Is that planned for? Phones or other hardware will be super-valuable compared to everything around them – and super different – so just like any other hotcake, they’ll get snapped up quickly. What percentage of product loss are you planning for in Week 1, Month 1, Year 1
- How will they be fixed when they break? How will users get a new screen, or a new case? Because they aren’t part of the ecosystem the repair facilities etc. don’t always exist.
- How will you make sure devices are charged? People will keep their personal device (i.e. the one they already have, before they get a free one) charged more than the free device, most of the time. – adding chargers etc. adds many layers of complication to a system – so suddenly you are creating a whole supply chain around this device.
- What infrastructure do they connect to for information flow? Internet? with what right now? where are the wifi hotspots? Or, do we need to bring in something like a VSAT. How much does that cost, per user, per megabyte? if there is no internet, you might as well use paper. Will they connect through sim cards? MNOs are already suffering from overloaded networks. Who will pay for the sim cards? For how long?
- Who trains people in how to use them? anything with training becomes many times more complex. I’ve just switched from iPhone to Android and it is taking me weeks to figure things out. How much training is required per person, per device? Who’s paying for that? The partners who you are giving phones to? Have you seen their training budgets first?
- How does this affect the local economy? What is the market distortion on local hardware sellers? How many vendors might get put out of business. how do you proactively make sure that you don’t break the country’s growing tech capacity? Who has done the business plan for this part of things?
- How do you make sure you don’t create a false incentive for the future? in some countries, giving hardware has created a future dependency where health workers will not be trained in any new program unless they are given new handsets.
Some further thoughts
The current infrastructure is weak
Mobile networks have limited data capacity. External solutions (like VSATs) are inextricably linked to external networks (which create financial dependencies). There is also not a lot of electricity. Big devices need more juice. Solar chargers, or any type of generation of energy solution adds another layer of complexity and dependency on an already thinly stretched distribution and logistics proposal. It is also difficult to move generators into parts of a country where there is frequently (or consistently) no road access. It’s even harder to keep those generators there and running.
Governments have to be involved.
Systems should be built that support existing infrastructure, and that build local talent and capacity in the governments of countries where work (for example, the Ebola response) is ongoing.
Getting fast data is a prime challenge. However – when we are building systems to address that challenge we have to build those systems at the point of need, and look at the infrastructure in which we are creating them.
Hardware is only a small part of any solution
Increasingly, people like health workers, teachers, and other frontline professional can get access to a phone that is pretty OK. They may not have data – because the networks are thin. But the problem is not lack of devices. It is lack of information.
There is a big problem with lack of information – and lack of data – but you don’t get that by distributing hardware, by itself. You get data, and information, by working with what people are already doing, and making it better.
Facebook knows this – that’s why it knows more about you than you know about yourself – because it’s made things you want to do easier – and every time you click it gets smarter. But it’s made your life easier, it’s made a complicated task more simple, and it gathers information from that to understand the world. It hasn’t asked you to fill out a complex survey on everything that you’re interested in – Facebook knows that doing that would lose your interest, and lead to much less useful information.
That’s how we should deal with health systems. Make life easier for nurses, health workers, and others – let them order supplies more quickly, or tell us about new and get a response – and then we’ll get all the data we need. This is a very different to assuming that simply creating a paper survey on an electronic form will solve the problem of lack of data. This may help ameliorate parts of the broken health system but it won’t move them forward enough to provide truly realtime national data for decision making.
The Ebola outbreak has brought out a tremendous amount of good sentiment and good-will in the technology community. It is vital that we channel that energy into things that will create strong and resilient structures over time, and that we work with local technologists and entrepreneurs in the countries affected by Ebola to ensure that we are helping them rather than simply doing what we think might be best from far away.
New York, 17 November, 2014
With input from Erica Kochi (@ericakochi) and Evan Wheeler (@evmw)