Big data for small businesses
With a relatively small team, Hydra Consulting needed a solution that would make the social media analysis easy, and would not require significant investment in infrastructure or training.
“We do a lot of work with start-ups and small businesses, so we were keen to find a solution that we could not only use for our own projects, but also recommend to our clients,” says Dr. Oemcke. “When we saw IBM® Watson Analytics™ for Social Media, it immediately fit the bill. We can use it to cut into the soul of an advertising campaign in a couple of days, rather than the weeks of work it would take to trawl through data using traditional methods. That means we can offer an engagement analytics service that even our smallest clients can afford.”
The Hydra team proved the value of the IBM solution by using it to analyze the social media impact of the Australian drinks company’s TV ad, compared to an ad broadcast by an American beverage company during the same sporting event.
Dr. Oemcke explains: “We used the solution to analyze social reactions about both ads in the days and weeks after the broadcast. For the American beverage ad, sentiment analysis showed that around 35 percent of viewers reacted positively, and around 25 percent were negative—which shows that people actually cared about the ad one way or another. By contrast, with the Australian ad, we saw much less emotional engagement: 77 percent of the people who reacted on social media had no emotional engagement either way.”
Next, the Hydra team used the solution’s keyword segmentation capability to find out which aspects of the ad were sparking the most conversation, and integrated data from Google Trends to measure the impact on search terms.
“Most of the ad centers around two main characters,” explains Dr. Oemcke. “But there is also an Australian model who appears during the last few seconds, and she retweeted the ad on Twitter. In the following days, there was twice as big an increase in the number of searches for her name, compared to searches for the advertiser. She also got a large number of reactions, while the company’s social media accounts didn’t see much of an increase in activity.”
Designing more effective campaigns
In the case of the drinks company, Hydra’s analysis revealed a number of lessons that advertisers could learn from.
Dr. Oemcke says: “Watson Analytics for Social Media not only helps us measure viewer engagement—it also shows us what the viewers are engaging with. That’s incredibly valuable for brand-building, because you can see what works and what doesn’t, instantly. So you can change course and refine your campaigns to keep your ads on-trend.
“In this particular case, a smart move might be to reduce the emphasis on the two main characters from the first ad, and give the model a leading role in future campaigns, or bring her on board as a brand ambassador.”
Looking to the future, the Hydra team is excited by the possibility of combining cognitive analytics with marketing automation to shape and pivot campaigns more quickly, and to interact more effectively with groups of customers, and even individuals.
“Cognitive technologies like Watson Analytics will drive the next phase of evolution in the way we plan, execute and analyze advertising,” says Dr. Oemcke. “In the old days, success was measured by impressions—how many people might have seen your ad. When online advertising took off, we moved towards the concept of clickthroughs and likes—how many people had seen your ad and followed it to visit your site or buy a product.
“The problem with clickthroughs is that they’re just about the immediate numbers: they don’t tell you who is excited about your ad, or who is falling in love with your brand. That’s the difference between a short-term sales bump and a set of loyal customers who’ll buy your products for years to come.
“Watson Analytics helps us translate the real, long-term value of advertising—building a brand that people love—into a language that the new generation of data-driven analysts can understand. By quantifying the value of previously intangible aspects of a campaign, we can provide a much fuller, richer view of whether ads are truly effective.
“The next step is automation, which is where technologies like IBM Campaign come in. In the future, we’d love to help our clients build campaigns that automatically analyze the sentiment of responses from different demographics, and push different content to each group depending on their reaction. That’s the future that cognitive analytics can unlock for us.”
He concludes: “Our first project with Watson Analytics for Social Media has been a resounding success, and we’ve been blown away by the support that IBM has given us. They really understood our vision of bringing big data to small businesses, and we’re looking forward to continuing to work with them to bring these advanced analytics technologies to our client-base.”