Thought Leaders in Mobile and Social: Interview with Kristin Muhlner, CEO of newBrandAnalytics (Part 2)

Sramana Mitra: To deliver on this particular use case, you are obviously pulling data from Twitter, Facebook and other social media channels. How does that data framework work? Do you have to pay Twitter to get their data?

Kristin Muhlner: No, we don’t. Twitter and Facebook are open feedback channels. As long as you comply with their terms of service and their developer APIs, there are a number of methods with which you can gather data. With Twitter, for example, we are listening for those specific handles, and then we can bring that data in. It works similarly with Facebook – there are pages on Facebook dedicated to the agencies where you can gather data.

SM: In terms of the analytics infrastructure, what volume of data are we talking about?

KM: It really depends on the customer. Some of our customers have tens of thousands of reviews or bits of feedback a month. Others make a much smaller number. We take all of that content, and as I indicated, we have a natural language processing stack that looks at each of the comments for overall sentiment and then composites them into its constituent parts. Continuing to use the DC example, I might post on Facebook saying, “Surprisingly good experience at the DMV today. 35 minutes to get my driver’s license. It was accurate and Marie at the front desk was very helpful at getting me through the process.” Through that post, we have learned independent things about the wait time, the delivery of the service, and about a particular service delivery person.

We take each of those items and our [software] composites that comment into its constituent parts and associates it with its appropriate category, grades it for sentiment and that is presented back to the agency head. We have a great deal of scale in our architecture to support many reviews in order for them to be able to get value out of it. What is interesting about this kind of technology is being able to take thousands of comments, boil them down to their meaningful scenes and the sentiment associated with those scenes so they can be turned into actionable insights. Without that, it becomes difficult for someone to find a signal and a noise.

SM: Let’s do a few other use cases.

KM: Let’s talk about a food and beverage example. One of our customers is [the restaurant] Ruby Tuesday. It is an interesting company that has had great success over the years and has gone through a series of strategic changes. They are now in the process of refreshing their brand and making changes in how they approach the market, both in terms of their menu and their pricing structure. Our partnership with Ruby Tuesday has been really a win for them in terms of helping them understand what the conversation is about the brand online but also about specific locations. One of the things we do is we listen to feedback about Ruby Tuesday in general, but we also look at feedback specific to particular locations. If someone walks out of the Ruby Tuesday Knoxville Airport and tweets about that experience, we will geo-sense that tweet to that location so we know exactly where it came from.

In the case of an organization that has hundreds of thousands of locations, like Ruby Tuesday, not only are they getting specifics on what customers are saying about specifically named menu items or the general experience, but they are also getting specific feedback about a particular location or operation. This helps Ruby Tuesday in a variety of ways. It helps them refine their training approach, for example. Their trainers use the data and maybe see that there is a preponderance of commentaries about silverware replenishments on the tables or an opportunity to do drink resales more quickly.

Those are examples of things they can use throughout their training programs to make sure that internal folks are complying with their policies and expectations. The data also helps guide them on changes to their menu items. Maybe they will take a certain menu off, for example. They can listen to reactions very closely and see if there is a preponderance of negative comments about that menu item being removed or track feedbackson new menus.

One of the other things we can do is look at the specific mentions where a guest indicates a willingness to return or to recommend a restaurant, and at the same time they may be talking about an experience or a new menu item. What that shows is that there is a motivating impact a certain menu item has, and as a result it is something that they want to try to promote.

But not all is in form of macro views on training or menus; it might also inform about pricing and potential changes – areas where they need to rethink pricing – or specific restaurants and regions. They can use the data to compare the performance from one location to another: “How is my Knoxville location compared to my national location or compared to my Atlanta location?” Looking across all of the operational categories they care about – general experience, food quality, service, facilities, decoration, parking facilities, etc. – all the things a guest may comment on, allowing them to measure those and look for areas where they need to improve or pat someone on the back for a job well done.

In many ways it enhances or enriches existing performance management tools that may be in place, existing business intelligence and marketing tools as well as social marketing tools. What we now do is give organizations like Ruby Tuesday a near-perfect understanding of the conversation about them online. We know exactly what people are saying and we know exactly which conversations are driving loyalty. This creates an 80-20 opportunity for them to focus on things that are most important and to build campaigns around those things. If they find that people love the complimentary biscuits that show up on the table, that is an interesting thing to turn into a Twitter campaign. We can also tell them that their Facebook followers are twice as likely to recommend them as their Twitter followers. That creates an opportunity to generate more loyalty and put the right messages in place at this particular venue online, so that they are not wasting valuable resources in places they shouldn’t.

This segment is part 2 in the series : Thought Leaders in Mobile and Social: Interview with Kristin Muhlner, CEO of newBrandAnalytics
1 2 3 4 5