Interview: Exploring Data Science at Brandwatch with Hamish Morgan
By Olivia SwainAug 23
As part of my job as a research analyst, I monitor social media closely throughout the day.
I have TweetDeck perpetually open on my second monitor with streams of important hashtags and influencers. What would you guess dominates 95 percent of these tweets? Links to articles, news about a company, or other “push” marketing messages.
In my opinion, and other reports, this is not how consumers want to use Twitter. They want to have two-way conversations. If they ask a question, increasingly they expect to get an answer – fast.
Recently, I commenced an experiment to see which of the nation’s top brands actually respond when asked a question on Twitter.
It’s not feasible to expect them to respond to everything – some receive thousands of messages each day – but they should have a system for picking out the most important messages and responding to those in real time.
This is done through technology that automates message filtering, prioritizing and routing (could do with something like Brandwatch, eh? – Ed).
My goal was to figure out whether brands utilize these kind of technologies, or if they have another method for finding important messages, or messages with high purchase intent, risk of switching brands or negative or positive feedback that could go viral.
The infographic below shows the percent of total tweets we sent that received a response, and the average speed at which each brand responded. We also summarized a few takeaways we learned from the experiment that companies can use to improve their own response.
The questions for the race fell into one of five categories and were designed specifically to prompt a response. These categories included:
Me and three other employees from my company sent the messages every weekday for four weeks from several different personal Twitter accounts.
Half of the time we used the @ symbol and the brand’s Twitter handle. The other half of the time we just mentioned the brand.
Here are some key lessons we learned from the experiment:
1/ Keep the customer informed
Coca-Cola and McDonald’s committed big errors when two of their replies came several days after the questions were sent. For the instant-gratification customer, this is the same as not responding at all.
2/ Don’t be a robot
Customer service expert, best-selling author and speaker Micah Solomon told me recently that being human in your engagements with customers on Twitter is one of the most important considerations. Twitter is a social platform, your responders need to talk and act like they would interact with their real friends and family. Say thank you. Be personal.
3/ Important keyword triggers are your friend
When we designed questions for the race, we specifically included questions with important intent, sentiment or risk of switching brands.
Social listening software can be programmed to send service messages to the front of the line if they contain keywords such as “help,” “mad,” “thank you.”
These rules are imperative for brands that need to automate tweet prioritization.
4/ Listen for your brand, @ or no @
The social customer service innovators watch and respond to non-@ mentions because they see the opportunity to really surprise and delight.
Most social listening software can be programmed to listen for mentions without the @, with the @, and #brandname.
5/ Still Not the Norm
These brands responded to a mere 14% of the 280 tweets delivered during the race. Whether the issue is one of strategy or technology, brands are still far from meeting customers’ expectations on Twitter.
Sign-up to receive the latest insights into online trends.