Campaign poster for the presidential elections in Liberia, October 2011.

Liberians will go to the polls on 11 October 2011 to vote for a new house of representative, a new senate and – most importantly – they’ll decide who will be the president for the next six years.  Ushahidi has set up a website to monitor the elections.

This is going to be the first time I’m in a country where Ushahidi is used to monitor elections. And while I’m skeptical about its usefulness as a tool for (first) responders in an emergency situation, I think that it can be a very useful tool to monitor elections or in development contexts.

I’m looking forward to seeing what reports Ushahidi will have from Nimba County and Sanniquellie, where I’m based, because I think that in Liberia Ushahidi has to deal with a number of challenges in order to be able to give a comprehensive picture:

• Many people don’t have mobile phones or no credit or no electricity to charge their phones. This has an impact on how many people will be able to send reports to Ushahidi
• SMS are not widely used by Liberians
• Very, very, very few people have access to the internet
• The monitoring project has not been promoted very well. I only found out about this weekend – from an ICT-interested expatriate.

All of this limits the size of the crowd that Ushahidi can draw on and I would be interested to see a demographic breakdown of the people who submit reports.

Update 11 October 2011: It turns out that Ushahidi is working with 11 organizations in Liberia who are submitting reports to the web platform. These include local organizations, organizations specialised on monitoring elections and UNMIL. While this limits the size of the crowd it probably means that the reports are more accurate and more topical. Particularly the inclusion of UNMIL is interesting because UNMIL is in charge of providing security in the country in case anything goes wrong.

Update 13 October 2011: Please read my follow-up post on how I experienced Ushahidi’s performance.

Related posts:

1. Web based election monitoring in Liberia: a failure
2. Use of technology in the 2011 elections in Liberia
3. How the BBC is trying to avoid the gullibility trap

The difference between data and information

Data or information?

In my experience one of the biggest problems is that people don’t understand the difference between collecting data and analysing information. When you talk to a beneficiary, fill out a form and then archive that form somewhere, then you have collected data. However, that data only turns into useful information once you have entered it into a database or spreadsheet so that you can filter, sort, aggregate or calculate data.

If you work for a big retailer, you might think that this is obvious. However, in the humanitarian aid industry it sadly is not.

Three examples:

1. An organization working with refugees announced during a meeting that they would relocate between 2,000 and 4,000 people from host communities to a refugee camp during the next few weeks. Another organization that is in charge of providing food to that camp, said that they’d need to have a more precise number so that they know how many sacks of rice to buy. However, in the rush of the crisis the first organization had only registered heads of household and was not sure how many people were part of each household.
2. Different meeting, different actors, similar issue: Somebody announced that 458 people would be relocated to a refugee camp in the coming week. Someone working for a an education agency asked, how many of these would be children for whom schooling would have to be provided. The first organization was not able to provide that information since, while they had captured the age during registration on paper, they had only entered “minor” or “adult” into their database.
3. A while age I got to work with a big Excel spreadsheet that contained data of people who had been registered. Many of these people had two or three names (Last name, Middle name, First name), all of which were entered into a single cell called “Name” in the spreadsheet. However, people would sometimes give their last name first, and another time say their first name first, making it difficult to find them in the list. This problem was made worse by the fact that many of the people were illiterate so spellings might differ slightly from one time to the next. Once we put each name into a separate field, it was much easier to find even the ones that were misspelled because at least one of the three names would normally be consistently spelled and then you could simply scan the list of remaining names for phonetic matches.

Granularity and structure of information

I realize that collecting data during an emergency is a very challenging task. However, the three issues I described are not about an inability to collect data, they are about a lack of knowledge how to treat data properly so that you can use it to inform your decisions. In other words: they are examples of bad information management.

One frequent answer to this is “True, but we didn’t think about that in the beginning.” Well, that is exactly why you need to have professional, experienced information managers!

Particularly during the first phase of an emergency, organizations should invest in people who can help them structure the information they need in such a way that it is collected and entered in a useful way. This is something that can even be done by consultants on short-term contracts. Then, once the structure is in place, other (less expensive) people can enter data into those spreadsheets or databases and programme people can use the information as a tool.

Data confidence and errors

The second big challenge is to ensure that the data is entered correctly and stays correct. Everyone makes mistakes, particularly when you have a mind-boggling boring job like entering a list of thousands of names or dates of birth into a spreadsheet. One of the most efficient ways to prevent data entry errors is the double entry method. This means that everything is entered twice, by different people. Once all the data is entered, you can automatically compare the two data sets and look at the differences. I know, I know … I can almost hear cash-strapped managers groan at this suggestion, but hiring two data entry clerks might be only marginally more expensive then hiring one more qualified person.

The second way to reduce errors and increase data confidence is to have programme people work with the data themselves. I’m not necessarily talking about the initial massive data entry, but of the subsequent steps. In a lot of situations, updates are simply given to a person, whose job it is to update the database. This person normally has no knowledge of the programmes or the cases and most of the time never leaves the office. So for him, the data has no context and he will happily enter things that are written down, even if they don’t make sense. On the other hand, a programme person will notice inconsistencies or even be able to suggest improvements to the database. However, that means that organizations have to invest into teaching their programme people the basics of information management so that they understand how to do things and why certain things are done a certain way.

Stop reinventing the wheel

Last but not least, I’m still amazed to see people reinvent the wheel over and over again. A lot of the data that has to be captured is the same in different emergencies, with only small adjustments being necessary to fit the specific context. However, in each emergency people seem to start from scratch and build their databases and spreadsheets as they go.

It should be the role of the global cluster leads to provide standardized information management tools that can be reused and adapted. And while some clusters are already doing this, it is still too few and in the cases where the clusters do provide these tools, a lot of organizations are not aware of them – which again calls for an experienced information manager who knows where to find these tools and how to use them.

I know that hiring people with laptops and spreadsheets is not as sexy as buying trucks or as photogenic as handing out food to beneficiaries. But I am convinced that if we are serious about being efficient and about using the money that we have been given to the greatest advantage of the beneficiaries, we need to professionalize and invest into information management. The people who can do that job already exist, you just need to have to the will to find the budget and hire them. You programmes will improve because of it.

Related posts:

1. From collaborative “crisis mapping” to “crisis feeding”

Not far from where I live, an organization that shall remain nameless installed two massive water tanks, which are supposed to serve a community. And to make things easier, this organization installed an electric pump that uses solar power. Unfortunately, what might have sounded like a great idea is completely useless in reality, because the solar panels never made it through their first rainy season and there is nobody who could repair or maintain them – today the tanks are completely dry and empty.

So yes, technology is great and can be sexy, but it has to be appropriate to the context and you have to make sure that you train people in their use so that they can continue to use and maintain it.

Related posts:

1. Use of technology in the 2011 elections in Liberia
2. Web based election monitoring in Liberia: a failure
3. Rural Liberia: Where mobile phones are still a rarity

In many villages, satellite phones are the only way to make a call - and even that doesn't always work.

This is my first time in Africa. However, the one thing that all my colleagues with Africa experience had told me was: “Everybody has a mobile phone.” This made sense to me based on my experience in Haiti where, even though the country is extremely poor, many people even had two mobile phones, one for each network. In Liberia – not so much.

Only 4 per cent of women have  access to a cell phone

Internews recently published a study showing that only 37 per cent of all male and only 4 per cent of all female refugees from Ivory Coast have access to a functioning mobile phone. Not “own one” but “access”! While you might argue that the refugees are probably in a worse situation than many Liberians, my feeling is that cell phone penetration among Liberians in rural Nimba is not much higher, either. (Obviously the situation is different in the capital Monrovia).

The main reasons:

• Many villages are not covered by the two big Liberian cell phone providers, Lonestar and Cellcom. Ironically villages that are right at the border sometimes fare better, because they are covered by MTN and Orange from Ivory Coast.
• The Ivorian phone providers block phones for outgoing calls if that phone has not been used for a while.  While this is also standard in a lot of Western countries – normally after one year of inactivity – apparently MTN and Orange block your phone if you haven’t used it to make an outgoing call for only a few weeks! And of course, the only way to reactivate it is to buy extra credit, which many people can’t afford.
• Some villages don’t have electricity, not even a generator, so people cannot charge their phones.
• In the villages where you have generators, you have to pay to charge your phone. This means that many people only switch on their phone when they want to call someone or when expecting a call.

Considering that many villages are in very remote areas and extremely hard to reach, this lack of phone coverage has direct implications for the delivery of aid. While in other countries, you can use mobile phones to inform communities about planned activities, in Liberia organizations sometimes have to find people from a certain village during market day in the next town, to pass a message to their community. While this tends to work quite well for news affecting the whole community, it can be more difficult when you want to reach particular people, especially refugees who might not be widely known in the host community.

 Sorry – moved houses

Take my work, for example: On a daily basis I’m travelling around the county to see children who have lost touch with their parents and for whom we have information regarding  their parents’ whereabouts. Sometimes this is just a short letter saying “Dear daughter, we are well. We are back in the village.” Sometimes it is a request to bring the child back to the parents (a service that the ICRC provides in some cases). Sometimes it can also be sad news, for example when a family member has died. The point is, we actually have to find and talk to the recipient of the message in person.

But a lot of times, when we arrive in the village where the child is supposed to be, he or she is either working on a farm (which can be a two or three hours march from the village) or we just find out that they recently moved to another community. But without any means of communication, there is no way to relay that information to us in time. As you can imagine, this can be quite frustrating and limits how many children we can see in a day.

“Mommy, I’m fine and I’m in Liberia”

Of course this also has implications beyond being an inconvenience for me and my colleagues: without means of communication, many refugees have not been able to tell their relatives where they are and that they are alive. When we go to the villages, we always have mobile phones and a satellite phone with us (whatever works) that people can use to call their relatives. So many months after their flight, I would have thought that the majority of refugees would already have had some way to get in touch with their relatives. But in a lot of villages that is clearly not the case and as you can imagine it can b very distressing not to know where members of your family are or whether they are even alive.

From what I hear from my colleagues, Liberia is the exception among many African countries and cell phone service here is significantly worse than in other countries. But is shows that mobile phones are not the one-size-fits-all solution that some people suggest it is.

Related posts:

1. Web based election monitoring in Liberia: a failure
2. Lack of communication fuels panic during riot
3. Liberia: Ushahidi to monitor elections

Anyhow, this project gives me the rather enviable opportunity to create something from scratch without having to worry about integrating any other systems or databases. And of course, in my mind I have played the “what if”-game many times. Now is my chance to put it into practice.

The conditions

• Website must be easy to maintain
• Must have a backend
• Must have a backend in a language that can be used by the people maintaining it
• Multilingual support
• Cheap
• Site should be up as quickly as possible
• Big developer community

With all that in mind I suggested to use WordPress as a CMS to drive the website. The National Society accepted this proposal and I have been fine tuning it since then. It’s running on a dedicated virtual server. For the design we decided to modify a premium theme.

Below is the set-up that I have in place so far. Please chip in, if you have any additional suggestions. I hope others will find it useful as well. All plug-ins can be downloaded from http://wordpress.org/extend/plugins/

• After the deadline
  Checks spelling, style and grammar of your English language posts.
  
• Category header content
  Let’s you add HTML at the top of your category or tag-pages. In my opinion, you shouldn’t need a plugin to do that, but it’s the only way I found to add custom text to the top of category or tag-pages.
  
• Contact Form 7
  A customizable contact form. Supports many languages.
  
• Dagon Design Sitemap Generator
  Generates a human readable sitemap.
  
• Easily navigate pages on dashboard
  The site will have a lot of pages (as opposed to posts). This plugin makes it easier to jump to individual pages.
  
• Google XML Sitemaps
  Adds a Google XML site map.
  
• Lightview Plus
  Improves the standard WordPress gallery and has slide show features. I have not figured out how to make this multi lingual yet. The plugin requires that you buy  the lightview script for 3 euros.
  
• Media Tags
  Allows you to assign tags to your media files which will make it easier to find them in the future.
  
• Registered Users only
  Hides the site from anonymous users while still under development.
  
• Sociable
  Adds Twitter/Facebook etc. “share” buttons to all posts and pages. I love how customizable this plugin is!
  
• WP DB backup
  Emails me a complete database backup every day. The frequency is customizable, as are the tables that are being backed up.
  
• WPML multilingual CMS
  This plugin is seriously impressive! WPML does not only create a multilingual structure for your site but also helps your editors with the translation workflow and even supports translations of widgets and text-strings (though that doesn’t always work 100%). Since my day job is to maintain a multilingual site I can tell you that I’d be much happier if we had this!   The plugin was developed by a translation company that integrates their translation services into the plugin. Very smart.
  
• WPtouch iPhone theme
  Delivers site news for mobile devices (not just iPhones). Since the site is for a disaster-prone country where mobile phone are widely used, this could be a useful feature.

I have not installed any YouTube and Flickr plugins yet because the one’s I’ve seen haven’t really convinced me.

I am still looking for a good comments-plugin that is available in multiple languages. Please leave a comment if you know one.

What would you add to this list?

No related posts.

Communications is only a small part of that response and social media an even smaller part. Nevertheless – here are my observations:

Convergence is already happening

As soon as the extend of the destruction became clear, American Red Cross asked the public to donate 10 USD through text messages for the Haiti response. Within the first day AmCross collected 800,000 USD. After six days they had collected 21 million USD. Since the appeal was not only spread through social media but also through mass media, it is difficult to measure how big a part social media played. But I think that the effect was significant. Because unlike when seeing the message on tv or reading it in the paper, many users didn’t have to switch device to take action.

What I mean is this: Since many people in the US use Twitter on their mobile phones, and since the donations happened through text messages, very little effort was needed on their part. They received the call to action on the same device they needed to take action.

Ease of use taps donors’ wallets

Other Red Cross and Red Crescent Societies had similar text messaging programmes. But while these also raised money, none of them were as successful as AmCross’s. I’m convinced that the reason is primarily that mobile phone technology and internet use have converged more in the US than in other countries. (I’d be really interested in insights from Japan on this point)

NGOs and non-profits should take note of this development and design a mobile phone strategy as soon as possible, no matter where they are. (see also: “Time to get mobile“)  In the US it is already necessary, because donors will soon expect this level of ease of use when making a donation. And outside the US, organizations have a chance to be slightly ahead of the curve when convergence comes to their country.

Content rules

Investing in photography and videos pays off. The public and the media have an immense hunger of exclusive footage from the ground. In the first few days quality is not that important, but that quickly changes and the higher the quality to start with, the better. Because we had good content, we were able to pitch our photos to media and got noticed online.

1 million views on Flickr

Not the best photo in the world - but seen 70,000 times.

All photos in our Flickr set about the Haiti Earthquake combined generated over 1 million page views within 24 hours on January 14^th. It was highlighted by Yahoo! (which contributed the majority of impressions) but other media paid attention as well. In the first few days, BBC Online mentioned it on their live blog every time, we uploaded new images. We also got a substantial number of requests from media who wanted high-resolution versions of our Flickr photos.

Return on investment

As far as ROI is concerned I should mention that most visitors stayed within that set and did not click on other photos or through to our site. So while this was very successful to generate awareness, it did not generate substantial funds for us. However, since Flickr’s community guidelines  forbid actively asking for donations, there was no call to action under these pictures either – merely “Find out more at http://www.ifrc.org/haiti/ ”. And besides, our role as a Secretariat is primarily to highlight the work of National Societies – so for us that still is a success.

Quick and easy tools to help spread the message

We’ve also made our Flickr set available as an embeddable slide show and share the code on Facebook and with National Societies. I have no information whether this is being used a lot, but since it only took two minutes to set up I think it was worth it.

CNNireport: From online to on-air

Finally, we uploaded the photos to CNNireport, CNN‘s “citizen journalist“ portal. While this did not generate a lot of views online, CNN used a lot of these images on air.

Video: be creative

Video was – and is – much more challenging. Not only are videos more difficult to produce, there are also bandwidth issues. While photos could be sent from Haiti by mobile phone, there simply was no bandwidth to send high quality videos in the first few days. However, AmCross showed that simply having someone on camera who can talk intelligently about the situation on the ground can be enough – even if that person is far away: Tracy Reines, director of international response operations, did short video messages in the first few days in which she explained what the Red Cross was doing. Her first video was seen more than 200,000 times on YouTube. Unfortunately there was also an incredible amount of extremely racist and obnoxious comments which makes me believe that it might make sense to pre-censor comments.

3. Facebook, Digg and Reddit

We routinely post new content to our Facebook page, to Reddit and to Digg. We have never been able to generate much attention for our content on either Reddit or Digg, Facebook however was a surprise to me. It was surprising to me how little impact it had. Our stories on Haiti got pretty much the same amount of “likes”, comments and shares that most of our day to day stories get. I would have expected much more. Something I didn’t do – and maybe that was a mistake – is create an album with photos on Facebook, similar to what we did on Flickr. Maybe that would have worked better.

Crowdsourced mapping for Haiti.

4. Crowdsourced mapping

I actually want to do a separate post about this topic, because I find the crowdsourced maps that are available about Haiti extremely impressive. We haven’t been actively involved in this ourselves, but I think we have to find a way to include these resources at an operational level. For the time being, please take a look at: http://haiti.ushahidi.com/

5. What did your organization do / learn?

These are my first thoughts and experiences from a social media perspective. I’m currently on my way to Panama to assist our regional office with supporting our teams in Haiti. This will be general communications support – not social media specific – and I’m sure I’ll be too busy to blog once we have landed. But I’d love to hear from you: what your organization has done or learned about social media in emergencies. And even if I don’t have time to write, I’ll find the time to approve comments. So please share your knowledge!

P.s.: Actually I’ve been to Panama for four days now – but didn’t get around to posting this before today. It’s great to see how the Red Cross Red Crescent is doing everything to help the people in Haiti. And it’s a real privilege to be part of that team.

Related posts:

1. From collaborative “crisis mapping” to “crisis feeding”
2. How the BBC is trying to avoid the gullibility trap