Thunderclap allows anyone to create a campaign that blasts social media messages from many users accounts all at once. We call each blast a “Thunderclap campaign”. Each Thunderclap campaign is designed to rally people to the campaign organizer’s target landing page at a crucial time. Every week, hundreds of new campaigns are created and hundreds of thousands of social media messages are sent to their supporters through Thunderclap.
InstigatorResearch + PrototypingRailsPostgresChartjs
For every positive email we got about the success of a user’s campaign, we got a confused or angry response questioning whether or not their blast went out at all, like the ones below:
"please refund, the campaign did not work. :("
"how do I know it's working?"
This was a trivial problem to solve. We created a confirmation email. When speaking to users, however, we discovered a related and more fundemental problem: campaign organizers had no idea how successful their campaign was. We may as well have been selling snake oil.
Doing some research to better understand this problem, we learned that many of our campaign organizers, particularly those who used paid campaigns, were being held accountable by their organizations to ensure the money they spent was worth it. Brands as well as NGOs were becoming increasingly metric- oriented and campaign organizers needed to show quantitative results.
We broke this problem down into three key components:
How We Solved It
First we had to re-evaluate what success meant. Was success just about the size of a campaign? With raw numbers, larger brands would always appear to be more successful, but there were plenty of upstarts that had viral campaigns that were massive, relative to their size. We had to figure out a way to score success relatively.
Another challenge we faced was figuring out the true scale of a campaign. Up until then we had focused on the number of backers who lended their social media accounts to the campaign, but what about retweets and shares and all the other ripple effects a campaign could have?
In response to these questions we came up with something we called the “Amp Score” (short for Amplification Score). The Amp Score is calculated using a campaigns’ page views, view-to-supporter conversion rate, and click-throughs from blasted messages to the intended landing page. I built a scraper that would search for conversations related to a user campaign’s hashtags as well as track the engagement on messages in terms of Facebook likes and Twitter favorites.
Once we figured out how to measure success, the next step was coming up with a way to communicate it. I was inspired by Des Traynor’s Jobs to be Done Framework example, where they built a robust, interactive map tool for customer support teams only to discover that marketers were taking snapshots of it to show their bosses at the next meeting. I pushed this idea of a ‘success artifact’ - an auto-generated dataviz graphic sent to the campaign organizer after the blast that can simultaneously congratulate them on a job well done and be shareable to their audience. I created a mock-up and we decided to test our theory by manually sending it out to a subset of users.
Users loved the success artifact and started sharing it with their supporters and so we built a feature that would auto-generate and send the artifact out after every campaign.
For users who needed more robust metrics, I revamped our campaign stats page to give campaign organizers more insight into how their campaign played out.
I added a map that captured new insights on where the campaign was most popular.
I udated the table full of supporters to be sortable across several several attributes:
I created an interactive stacked chart showing cumulative campaign growth, per social network, with the option to view in terms of total supporters, their social reach, or the amount of times they’ve shared their support to their timeline.
I created another interactive chart showing where most of their supporter traffic was coming from, and breaking down their views/conversion rate.
I added another section called Engagement Report that lists engagement on the messages that went out per network. It also now counts the second-wave ripple effect of engagement on your supporter’s messages and an analysis of the conversations happening around your hashtag, breaking down sentiment, related hashtags, and gender of your audience.
Now here is what our email complaints tagged as 'blast confusion' looks like since we rolled out the success artiface in the middle of May 2017, and the revamped stats page a month later
The effect of all this was a massive drop in confusion.