2020: Omnichannel presentation
How to relaunch digital content experience with no budget
During humanitarian emergencies like typhoons, earthquakes and disease outbreaks like Ebola and Zika, the World Health Organization (WHO) is there. It coordinates over international 900 partners to relieve suffering and save lives.
However, in 2017, WHO was managing a massive website of over 150,000 pages on a CMS that was no longer being supported on deprecated web servers that could no longer be upgraded. Digital channels were siloed. The organization’s digital ecosystem was no longer fit for purpose.
What did WHO do to turn the situation around? What strategies were used to break silos and move to a more modern, omnichannel approach?
OK. Thank you, everybody. I am now Sabena. I'm here with Chris Strebel from the World Health Organization. Chris is a day job is digital content experience, which is creating content enterprise content systems at the World Health Organization. This often involves providing digital and connection sports for outbreaks for crises like Koban 19, Zika and Ebola. So we could not have had a more timely presentation. We did not plan it that way. But it's it's it's kind of been a very interesting alignment of the stars.
Very happy to have Chris with us. He's an extremely busy man. As you can imagine. So we are thankful and honored for his presence. Fun fact about Chris outside of his work. He likes competing in triathlons and skiing, and he lives in Switzerland. So that's a great place for both. So without any further ado. Chris, please take it away. Thanks. No. Great to participate in this, hoping to do it in person, but I'm happy to present what we're doing at the World Health Organization, presenting on the mass web migration we're doing, which is taking it from a more static content management system to an enterprise content management system with all that entails and how the moving of one hundred eighty thousand pages and lots and lots of PDX is the easy part of what we were doing.
So in the past, the W8 show was the place to go when there were typhoons or floods or earthquakes. You can see a story here where we had a drop in cholera cases. This is in the olden days. And we also support things like as no said coronavirus. We were there for Ebola and Zika and were there basically to support the world's health and make the world a healthier place. So the story really starts in twenty eighteen, which is this is what our Web site looks like.
And one of the challenges or the challenge is really where there's a technological one. And that was basically we had the CNS itself was no longer being supported and the servers and systems were no longer being updated. So the servers were all and the security patches were no longer working. So from a technological standpoint, we needed to do something. You can see from a design standpoint, it was very static. It was very hard to change. A lot of content had to be updated in many, many places.
And so we would do a new story on the home page, the news page, the country. It came from page three or four other places. And if something needed to be changed or updated, somebody had to remember all that stuff. And that was not so easy. So it was it was time. And I'm just to give you an idea of. How old are you? This was this is the most popular technology of the year when we instituted our old C.M.A.
So, as you can see, really something that we needed to tackle and to make more dynamic. So I just put together a quick agenda to kind of let you know where I'm going. And so you can kind of follow a little bit better. I going to talk about how we got by and for this project, because we started with absolutely no budget. Some tips to migrate a massive Web sites in multiple countries and multiple languages. And then basically how to leverage stope create such a project.
So in terms of buy in.
It was really we had to justify it into senior management for years and years. We were saying, look, the system is breaking down. We don't know when it will break. But if it breaks all content, we'll be stuck in time for for three to six months or maybe more. You can imagine if that happened right away and coronavirus, our website wouldn't have updated, which would have been an absolute disaster and really needed some risk mitigation that that alone was hard enough to just sell the argument.
What really sold the argument was we made a return on investment argument where we went in looking at what other units across the house for investing in the Web. And we realized in two years, about two point seven million could be saved. So we got in an initial investment to start a proof of concept. In fact, four projects were presented with fifty thousand dollars. And of the four only the Web site part, we accepted the 50 and said, let's see what we can do.
Fifty thousand dollars rejected things as far too small, which probably, most definitely was. And so we accepted it and we got the other two hundred thousand. And then after the proof of concept, people were very excited and upper management about what we had done and that turned into several million. So we went from fifth from zero fifty thousand to several million. So my advice is to accept what you get, to prove what you can get with a with a proof of concept.
And basically, this is the next stage was our proof of concept. We couldn't do all one hundred and eighty thousand pages, but we focused on the newsroom highlighted in orange here and all the subcategories in the left hand column. And we had no mega menus at that time, so that was a big advance. We had no dynamic updating of the news in the middle of the mega menu. So that was something new. And we had a new layout and automated structure.
So we published news once that published in many places. So that was. That was really useful because managing the news part for me was a big part of my jobs that suddenly gave me a little bit more time to focus on migrating everything. Part of the strategy between me and choosing that particular part of the site. So another part of getting the Buy-In is getting it from the very top level.
So we went to the top level for our corporate headquarter sites to the director general, who a lot of you. I'm sure I've seen on television for the World Health Organization, Dr. Tedros. And he's a very positive guy. He loved the idea. And he said, yes, let's do it. We started. But then he didn't. Even though he was supportive, he didn't allocate it to the next level down in our Western Pacific office. What happened is the top guy also saw it and he told his next in command.
I want you to follow up on this project and make sure it happens. And so right away, that's a difference between an interactive top level Byan and a little bit passive. And the interactive is for us. What made the huge, huge difference on this? So we started with the Western Pacific. We migrated their whole Web site across. And these 36 other country offices. As soon as that happened, our Southeast Asia office called. And the top guys there were like, well, we want the same thing.
We did it for another trying to come to Europe and another 11 offices. And suddenly we had a good third of the world's population being covered by these Web sites. So that's really good.
Interest would potentially be bigger and our headquarter site, except with a number of recent outbreaks. And we've also had other regional offices like Europe and Africa, the Africa region, also vying for our attention. So we're going to have to see where we go next with all of this. As you can see, the ancient past was only 20, 19. It's not that long ago we were doing all this. Of course, when you're doing a major migration and something like COGAT 19 comes in, you've got to be very flexible and adaptable.
And really, one of the ways you get by and is by being useful and providing services to people that they wouldn't get otherwise. With your migration, just show them it can be done and not invest in outside technology. So that's something we did at the time. We we really focused on the covered nineteen outbreak disease. And you can see the this is our original home page at the time when Kovik came out so that we started in in January.
And this was because we had very large Schiro images all over the site and it made it very, very difficult for people to find the actual Crovitz an outbreak on pandemic information. As you can see, there is a little button down there on the bottom saying find out more. And we ran some heat maps on that and basically nobody was finding that.
So with COGAT 19, things started relatively slowly. We had some disease outbreak news in December. And in January, we were publishing our first full on new statements about it because we knew it was a novel coronavirus. Early on it was people weren't sure exactly what it was. And the news started to slowly come in and see 20th of January, 5th of February, and it starts to speed up. And it became more and more until finally it was just popping up every day and it was becoming all consuming.
So that's basically where we were next in. Migration and the corona virus. Here you can see basically what happened to traffic. We went up to several tens or even 100 million visitors a day. You can see it's a one thousand two hundred percent increase in traffic with one point to almost one point three billion page views from the beginning of the year. So it's massive, massive traffic. And as you expect, things kind of started to happen. That's our C.P.U and our load spiking because of all the traffic.
And obviously, this led to new and increased things that we needed to look at and make sure stayed up. We had a Verant vendor report analysis. These are the things you don't want to see. That C.P.U goes to 100 percent on all on all the machines and the load balance marks are unhealthy that you just don't want to see. So some of that was because we set up our original proof of concept as a small proof of concept. And in the ensuing year, we were building on more and more of the health topics and we were chunking content into small groups and moving it through the system.
And so the systems were not set up for a massive spike in traffic or in usage. And in fact, what happened was it was the the people uploading a lot of documents that was causing the system, the crash, the traffic.
We had a good system in place, but it took a lot of analysis to figure this out. And it also required us to start to move from the zero cloud, the W.S. Cloud to the zerk cloud. And that's mainly because we have expertise in-house on as you were. And we have more people who can jump in and fix the situation. Thankfully, we never had the server go down for a long period of time. What went down was the backend editing.
So as far as things go, that's you're starting to almost have a heart attack. We're saved at the last minute. We also looked at various heat maps and studies to improve the user experience during the Cobbett 19 crisis. And so here you can see three thousand eight hundred mobile use of of the front page of the Web site. This is when we went to our new design where we got rid of the hero image and we put in a lot of COGAT boxes in areas.
And you can see the donate button up at the top here. It's getting quite a lot of traffic, a lot of views. You can see the red box is getting lots and and down the whole table, you can see it's being used. So that's mobile. If we go to desktop, we see a similar a similar pattern where, again, we're the red areas where people really need to go to find out everything about the disease. We have some quick links there, which some have said are not so quick links, but we're doing some research on it with top tasks.
I'll show you a bit later to really take people to where they need to go. And then we have this interactive country map. We've had three or four competing iterations type something that's dynamic with numbers and data countries. And that's also working really, really well. As I mentioned, we were working on top tasks. This is with the customer care words and during the government has is doing some pro bono work with us. There are a lot of companies out there doing pro bono work side, really helping us out basically in this analysis to ask users to select the five most important things to them.
And you look at the consistency and you look how you can service those users. And here you can see that we have nine or 10 really top key tasks that people want to do normally in a Web site. You have fewer like, for instance, on a travel Web site. You might only have booked a flight or in a hotel book, a room. And that's where you focus on you focus on that upfront w.h show. It has so many tasks.
And we're really going to need to have more of a link structured navigation for people to get their. We've also looked at these associations that when we create these these menus for people, that they're associated with things that people look at. So we've had over 500 people look at these various tasks and group, for instance, this grouping together where we can see most people associate avoiding physical contact, cleaning and disinfecting waste disposal together, personal protective equipment, as is associated with those first two at fifty seven point fifty three percent, respectfully.
So that's another good one. You can see right down the table that apart from the last one, they all seem to fit in well. So we're organizing things in ways that users can find them better and trying to take it to a better place. The other thing I wanted to show is some of the lessons learned and big strategic shifts that we have planned going forward. So one of the things I've heard, I've already mentioned the the mega menus and caching.
Another thing we're working on is page size. And that really means getting things to a level where they're uploading quickly and we're using chunked content, which enables us to do things for apps and bots and go. And it's really pushing us to take an OmnichannelX approach on things. We can really push things out multiple channels at the same time with a consistent approach, because what we had in the past was one block from would update its information would be in one system, another clock from would update its content would be another system and a third one.
Same thing. And what happened was, let's say our guidance or our guidance on wearing masks, changes. Then you have to remember to update all three. Now it's only three things. So for things, you might think that's easy. Well, what happens if you have it in six languages? That's our our our standard of six languages. At the same time, our DG is saying to push it out to all the people need things on coronavirus.
And he's absolutely right. So now we've run something with Viber where we have it in twenty three languages. You can imagine the difficulty of dating in twenty three twenty four languages as it goes on. The other thing we were doing at the same time was onboarding governance, because as soon as you make an easier tool that doesn't necessarily make an organization smarter and it just keeps the organization same as it is for more people have access to do what they were doing.
So if one doesn't control that, we have to be very careful. We didn't end up with a mess. So the scalability and the governance and controlling that so that we didn't go off in all directions was super, super important.
I've talked a little bit about the Structured content, so we're structuring it from and breaking down our health topics and too many different areas so that they can be pulled off. It's being used in Google knowledge panels at the moment. If you go on to Google and you search coronavirus, we've been working hard with the team at Google to get things surface easily there and make it reusable so that if a team is working on a surveillance that can be used along with everybody else working on surveillance, I'll get to that a little bit more later.
So one of the big this is how we did the content chunking, we started it off on health topics where this was an earlier iteration of top tasks, where we looked at what are the top 10 tasks just on WHL and not on coronavirus. And we were able to surface eight out of 10 on the health topic page because most people want to know an overview of what's going on with that diseases is because you may not know what is a corona virus, how to treat it and what its prevalence.
And then then there are other people who are more of the experts who want to know what's the data on it? What are the latest guidelines on it? What are the medical terms and conditions on it? What are the fact sheets on it? So we have a lot there in the right hand call. You go past one of the other things that we did as. As part of this revamping of the Web site was again the chunking down of content that we focused on on a brand new way.
And that was through activities. And we didn't have this before. And this was part of our work with the Western Pacific region where people were saying they couldn't see if we had a conference on surveillance or malaria. People would just see that a lot of people are sitting around a conference room working on malaria, but it's very hard to have a landing page just for what was going on on surveillance, because it was all news centric and the news was this conference.
So instead, what we did was we took things that people wanted to know about what WHL was doing. This is mainly for the donor audience or the countries looking to see what you chose doing. And it contains everything on surveillance. So we'd have that conference. We would have all the other things that are happening around surveillance, all the other progress being made. So this is to show not only the citizens of the world, because we really care about what people think, what what we're doing, but also for the governments for giving us our money to say, yes, they are doing this work and they are moving in the right direction.
So this is one of the key elements and changes. And so by chunking the content into activities and by chunking it into into the health topics in smaller ways, we now have less of a need for teams pages. For the actual departments, it becomes more of a an outreach of the directors to the world to show what they oversee. And it doesn't have to be as in-depth and allows us not to have to migrate every every page of content. Some of the other things we're working on very hard as some thinking, more dynamic and interactive maps that, as I showed you early on, is actually we've gone through three different teams have been working on it, and now we have one that we settle on for the moment.
And we're trying to each time we do things, do it in a way that's sustainable and can be built upon in the future. This particular one here is part of our data system, and our data system can be used in multiple health topics, in multiple activities, and it can be plugged in. And people don't need to have a in-depth knowledge within W.H of how the data is being generated. As long as I pick the right widget, it gets plugged in there and.
Yeah, that's that's basically that's basically a worldwide view of how we've dealt with getting a lot of buying some tips for for migrating, migrating the contents and keeping things moving forward in a huge crisis. Thank you very much.