Manager of Business Integration and Analysis
Customer input becomes continuous improvement
Text analysis of visitor surveys allows Alberta Parks to respond to customer concerns during the season, instead of waiting for next year
Jared Prins remembers the casual conversation vividly.
He’d been digging into an issue that had been flagged in the surveys of visitors to Alberta Parks, which manages about 250 campgrounds and 14,000 campsites that receive more than 1.8 million overnight visitors every year. Many of the parks don’t have running water; visitors who had been surveyed felt strongly that those facilities needed hand sanitizer dispensers. Taking a break, the program analyst with the Parks Division of Alberta Tourism bumped into the department’s assistant deputy minister.
It’s been an evolutionary change in the way we do business, in terms of processing this kind of information and drawing insight from it.
“He asked me what I was working on,” Prins recalls. “I said, ‘I’m doing this analysis, and I came across this issue of hand sanitizers.’ Right away, he keyed in on that as something that was actionable. We could do something about it, and he was fully engaged.”
Prins was working on data from an Alberta Parks camper satisfaction survey, a key source of data for the department. It’s now called the Reserve Alberta Parks (RAP) survey; an invitation to complete it is automatically sent by email to customers as they leave the campsite. In the 2013 season, from May to mid-October, the department received 15,000 survey replies.
“Almost 80 percent of those people provided text-based data,” says Roy Finzel, Manager of Business Integration and Analysis for Alberta Parks. “When you get that number of surveys returned, that’s very significant.”
In the past, Prins would spend three weeks at the end of the season inputting the text data, manually assigning a code to each comment. Now Alberta Parks is using SAS® Text Miner, which applies information retrieval and data mining techniques across a variety of feedback channels – phone calls, email, surveys and social media – with both structured and unstructured data. Instead of waiting for an end-of-season slide show, Alberta Parks’ regional and district management get weekly feedback based on those text comments.
That feedback leads to midstream operational changes with an immediate impact on the customer experience, says Finzel. For example, using SAS Text Miner, Finzel and Prins found that park visitors felt the coin-operated showers didn’t last long enough to warrant the cost. In midseason, Alberta Parks adjusted the length of the shower based on those comments. (Ironically, later surveys showed customers didn’t expect a shower that long, and worried the department might be wasting water. The shower length was adjusted again.) Alberta Parks can respond throughout the season to customer feedback about everything from the cost of firewood to the timing of caretaking operations.
“That’s the dynamic of this,” says Finzel. “That’s what changes from an organizational or cultural perspective. We’re becoming more responsive.”
That’s helped drive extraordinary customer satisfaction numbers. Eighty-four percent of campers were satisfied with the quality of services and facilities, according to the 2013 survey; 89 percent would recommend the park they had visited to friends.
On the near horizon for Alberta Parks is the implementation of content categorization and sentiment analysis tools, with a particular eye to mining social media content. But the tools are not the be all and end all, says Finzel. There has to be a cultural commitment to applying the results operationally and strategically.
“It’s been an evolutionary change in the way we do business, in terms of processing this kind of information and drawing insight from it, and the response from the management staff in the various regions,” Finzel says.