Online and Offline Mystery Shops Perfect Search Engine Algorithm

Posted by TrendSource on May 09, 2017

Background and Client Objectives

A top global search engine created a fintech service that used proprietary algorithms to pair would-be borrowers with lenders advertising on their platform.

After providing income and demographic information, borrowers would see a custom ranking of lenders based on the Annual Percentage Rate (APR) and borrower fees they would charge. They could then connect to the lender via click-to-call.

The company, however, needed to ensure that the APR and borrower fees the lenders provided did indeed match the rate they quoted borrowers on the call. They further wanted to integrate a quality scoring system into their rankings, one that considered wait times, response rates, and overall borrower experience.  This would ensure that their tool was not only providing accurate information to their users but also a rich and rewarding experience.

Essentially, they needed to understand the complete user experience from searching the service all the way through the follow-up call with the lending institution.

Program Development and Methodology

TrendSource knew the best way to evaluate an online user experience was through online mystery shopping and developed a series of intricate scenarios Field Agents would need before secret shopping the lending search engine to trigger a match with particular lenders.

This included a fictitious annual income, employment history, lending history, and credit score. Once they entered this information into the website, they were paired with a lender who they would then call, and Field Agents rated the quality of the call and the accuracy of the financial information provided.


TrendSource executed 200 calls per month over the course of two years, and fed the data into the client’s reporting website, which they then used to refine their ranking algorithm. In addition to quantitative data related to the accuracy of the quoted APR and borrower fees, they received data related to hold times and call durations, and qualitative data related to the Field Agent’s overall user experience with the representative and financial institution.

This information helped them to build a unique quality score for each lender, which they could then merge with their APR rankings to create one combined metric to determine search ranking order when providing search results to potential borrowers.


As market researchers look for new ways to track customers in omnichannel environments, they must devise and refine methodologies that adapt to both online and offline experiences. Online and Offline Mystery Shops, in this particular case, was just that methodology. It provided the client with a third-party perspective that tracked the entire borrower experience, moving from online to over-the-phone, mimicking how an actual borrower would engage the service. No solution fits every need, but mystery shopping proved flexible and provided the client a reliable, complete data stream.

Download the PDF

Topics: Mystery Shopping, Technology, Finance, Fintech