It feels great to share success stories of your clients and partners - and even more so when these stories are shared by major media outlets. We also celebrated the team's win, and decided to interview the service's technical director and co-founder Alexander Krakovetsky.
YS: How did you come up with the idea of the project, and using YouScan for social media monitoring?
AK: DonorUA's donor recruiting system was launched in March 2015. A few months prior, our project won at the IDCEE 2014 social innovation hackathon and received a grant from "Focus" magazine.
The project's main goal is to create a unified coordination center for donors, recipients and blood donation centres in Ukraine, as well as instant communication between all the stakeholders. The communication channels were our main development focus. This project was created in colaboration with the Ukrainian Youth Social Organization "Ukrainian young donor association" represented by Associastion President Irina Slavinskaya, and IT company DevRain Solutions represented by director Alexander Krakovetsky. At the present moment, the team consists of more than 20 people.
After the launch, we quickly shifted our focus to social media, which is often used by people to express their desire to help or call for help in various life scenarios. There are hundreds of social media groups that call for help and urge people to become donors, which raised the question of how to quickly identify relevant groups and posts. It's basically impossible to work with this kind of post volume manually, so that's how YouScan came up.
Since I have experience developing and working with similar software, I knew that a social media monitoring tool would be the perfect solution for finding these types of posts. That's why I reached out to our colleagues at YouScan, who gladly agreed to help.
YS: Let's talk about the process itself. How do you look for posts, what kind of content are you looking for, what kind of challenges have you encountered along the way?
AK: We look for two types of posts: posts that express the desire to donate blood, and posts that are looking for donors. We also recently started monitoring mentions on the topic of corporate donations - we want to get to know the companies that have adopted this practice.
We've had good initial results with identifying posts that search for donors, so our number of donor requests has increased. Searching for donors proved to be a bit of a challenge, so we had to work more on that. We're continuing to experiment with different types of queries. The good news is that posts about wanting to donate can often be found as comments to posts with calls for donations, so that works out.
YS: How many posts do you find? How many staff members are dedicated to this kind of work? How are the posts processed?
AK: According to our statistics, we receive up to 10 target posts a day (we're only working within Ukraine). Right now we have one dedicated team member who selects the relevant posts, contacts the users and tells them about our system and its benefits. Users can decide for themselves whether they want to register. The majority of them agree to sign up.
As you can appreciate, the nature of our project doesn't give us much time to wait for more posts to accumulate, because the urgency of donations calls for a turnaround of several days, or sometimes even several hours. So the question isn't whether we have 10 or 20 posts (it's better not to have any at all), but the high impact of each post. This is the reason why we're looking at bots and further automation of this process - this way we can reduce reaction times to mere minutes or seconds.
YS: Bots are trendy and fascinating, but there are many precautions concerning their use. Have you thought about those?
АК: We have. Right now our biggest concern is for our messages not to look like spam, so we're developing new user interaction scenarios. We even launched a startup dedicated to this: udonors. The main idea is to optimally automate communication through social networks and messengers, especially with a heightened interest in the creation of various bots. A little while back we won at the Startup Weekend Social Innovations 2016 in Stockholm with this idea. We're currently working on implementation and testing it out.
By the way, the first step has been completed: You can join our Telegram channel or follow us on Twitter, which automatically sends out new messages from the website. These channels are much more convenient for some people than text messages or email.
YS: All that's left for us to do is to congratulate you once again on the project and ask about your plans for the future. What are you developing next? How can YouScan help?
АК: Thank you, of course, we have some plans for the future. We'd like to use modern natural language processing tools (NLP) to automatically extract data from messages - phone number, email, recipient's name and contact info, blood type, city, etc. - then add this information to the database, so then we can get in touch with people to see how we can assist them. We already have experience developing similar software, but for different use cases.
"Urgent! My friend, a father of two lovely daughters, needs a blood transfusion! A+ blood type, single donor (please DM first, this is for a direct transfusion), and whoever else can help (with any blood type) can donate to a blood bank (Heroes Avenue...)"
While everything is in this disassembled state, we're carefully examining this channel, watching the users' reactions, gathering analytics, while simultaneously developing general interaction mechanisms. But the devil's in the details, so we're trying to avoid using trendy tools just for the sake of using them, and make sure that we know all the nuances of the process before using it.
We have already identified several features we'd like to see in YouScan, so I think we'll share them with you soon. And I'd like to take the opportunity to thank the YouScan team, who is helping the donor movement in Ukraine in its own great way.