Choice Hotels and Text Analytics – A Q&A with Debbie Tsusaki


Share on LinkedIn

Welcome to the first in a series of conversations on customer experience management with thought leaders across a variety of sectors. This series will focus on the different approaches and practices companies have with regards to capturing the “voice of the customer”. We hope that you’ll find these conversations useful, and that they might spark dialogues within your own organization on how you can engage and listen to your customers better and use those insights to improve your product/service offerings and operations to avoid crises and perhaps even identify new opportunities for generating revenues. Our first conversation is with Debbie Tsusaki, Sr. Director of Operations and Planning for Choice Hotels.

Sid:  Debbie, tell me a little bit about your role at Choice Hotels.

Debbie:  I’m a Senior Director for Choice Hotels and am responsible for any of the technology or applications used in our contact centers. My role touches on any tool that our people in the centers need in order to support our guests and franchisees – it can be processes, voice data applications, and the analytics which is where Clarabridge fits.

Sid:  So, who are the primary Clarabridge users at Choice?

Debbie:  The Customer Relations group, within the Choice Hotels Customer Care and Reservations department are, at least initially, the primary users of the Clarabridge technology. They process close to one million, text-based customer comments annually from hotel guests from guest relations case notes, emails, guest satisfaction surveys and other channels. The customer experience insight from the text analytics is used to provide feedback to the franchisees for help in making property improvements on a timely basis, by customer relations teams in coaching underperforming properties, and by marketing teams to understand the effectiveness of advertising and promotional campaigns, partner programs and overall brand performance.

Sid:  What sort of sources are you analyzing? Did you prioritize certain ones over others?

Debbie:  The first source that we prioritized was the guest comments, which can arrive verbally, in email or through a feedback form. These all get entered into our customer relations application. The second source is the comments made in the emailed Guest Insight Surveys. From the beginning, we knew those were the top two priorities.

Sid:  I know that you initially conducted a pilot program, which led to your selection of Clarabridge. What were you doing before you incorporated text mining into your customer care department?

Debbie:  It was a totally manual system. Back then, if we had a question about a specific property’s performance, we would have to go in to all those comments and read until we could identify the issue. As you can imagine, it was difficult. When we were determining the ROI of the Clarabridge solution, we found that it would take the customer relations team 15 minutes to read just five comments. With Clarabridge, we can now analyze thousands of comments about an individual property and then quickly create an action plan to improve that same property within just half an hour.

During the pilot, we continued business as usual – the manual effort – and evaluated Clarabridge against the status quo to ensure that we were confident in getting accurate information from the system.

Sid:  You’ve obviously found savings and efficiencies in automating that process. What else helped sell the value of the solution to you?

Debbie:  Automation is an important component, but also think about the ramifications of having not just automatic classification but consistent and accurate classification. Our human teams can’t keep up. The automatic classification of volumes and volumes of comments, day in day out, down to a detailed level means that our evaluation and recommendations become exponentially more detailed and actionable – more valuable because the [customer care] department and the franchisees know where and what action to take to improve their guests’ experiences.

Additionally, speeding the time from when a guest makes a comment to when the employee or general manager can see it, means that we correct and improve at a much faster rate. It’s hard to quantify the impact of preventing a problem from occurring, but it’s common sense that if we can respond quicker to one comment, it stops other guests from experiencing that same issue. That helps our loyalty and secures our brand.

Sid:  You’ve mentioned action plans that the team creates from the analyzed customer feedback. What’s the biggest change in the way you can analyze information now?

Debbie:  As I mentioned, it’s about being able to respond more quickly to guests, to have more consistency in data and a much finer level of detail. Our previous action plans were very high level. With Clarabridge, it’s very detailed.

Sid: For example?

Debbie:  In the old way, we could report that there was an issue with “Cleanliness in lobby”. That could mean a lot of different things, which is why when Clarabridge reveals “There’s dirty carpet in the lobby”, the hotel knows exactly what to fix to improve their ratings.

We have programs in place to help properties whose scores place them in the bottom quartile of ratings. The Customer Care group can now coach those properties with a much deeper and accurate view of the root causes of low ratings. With the ability to slice and dice the data in as many ways as we can think of, we can easily prioritize the biggest drivers of loyalty down to the individual property level. A property in a business center has different drivers than a tourist location, and knowing more than just “yes” or “no” responses to survey questions helps us nail down a positive experience for our guests. Comparisons can be run against a brand average, and the individuals at the property, our franchise owners and the corporate level have access to all of that customer insight.

Sid:  What advice can you offer other organizations that are considering using text mining software in their customer care organizations?

Debbie:  I highly recommend if you can’t do a pilot, start small. You’ll want to prove( cost justify) that you will receive real, concrete results just like we did. In selling this to our management what sold it was when we went in front of the executive team to pitch our business case. We proved through the pilot that we can realistically achieve savings beyond the cost of the project.

Second, be sure to go in with a very clear idea of what business problem you will solve by implementing Clarabridge. We wanted to reduce the amount of time it took us to research and create an action plan for a property that was experiencing issues. We wanted to improve consistency across our agents, from the way the data’s captured, to the category and sentiment models, to the way we report on the data. If you start manageable, and then keep adding to the project, you’ll be successful.

Sid:  We’ve seen that throughout all of our implementations – what else would you advise?

Debbie:  You have to have a subject matter expert on the team. You need the dedicated resource to ensure that the resulting application model and reports will be useful, day one, for the business owners. The SMEs have the tribal knowledge that will solve the business problem, I mentioned before. IT can’t do it, and bringing that expert into the process ensures a successful project.

Sid:  Were there any surprises?

Debbie:  Biggest surprise was how much time and effort we saved. We didn’t realize that it would be so significant.

Sid:  Music to my ears. Debbie thanks to you and to Choice Hotels for your time today.

Republished with author's permission from original post.

Sid Banerjee
A Greater Washington Ernst & Young Entrepreneur of the Year in IT services, Sid is the CEO and Co-Founder of Clarabridge™. Sid provides executive leadership and strategic direction and is a well-known expert in customer experience, business intelligence, and text mining.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here