Delivering qualitative market insights with Quirkos

delivering fashion

 

To build a well-designed, well-thought-out, and ultimately useful product, today’s technology companies must gain a deep understanding of the working mentality of people who will use that product. For Melody Truckload, a Los Angeles tech company focused on app-based freight logistics, this means intense market research and a focus on training sales agents as researchers.

 

Kody Kinzie, director of Melody’s special research and operations team, Cythlin Intelligence, was faced with introducing qualitative social research and analysis to people who had never considered themselves researchers before.

 

“Quirkos was the first truly accessible qualitative program I found,” Kinzie said.

 

Quirkos was designed with the philosophy that anyone can become a qualitative researcher. The goal is to allow companies and agencies to adopt unique ways to understand their staff and the wider marketplace. By making qualitative data visual and easy to code, users can see their results emerge and gain quick overviews of complex issues.

 

Companies like Melody are at the forefront of developing the next generation of qualitative insight, and Quirkos is helping to open the door to innovative new methods of business intelligence.

 

Kinzie started training his team members to use Quirkos but said he soon discovered that the simple coding allowed even a novice to develop complex data structures with notable uniqueness. Often, he found that these code structures were well suited to analyzing particular elements the researchers were interested in, and he began documenting the experiment to evaluate the resulting structures.

 

Here’s how Kinzie and his team use Quirkos:
One team member will send a Quirkos database to another team member — a researcher who examines the code structure and walks the requesting team member through an explanation of the thought process that went into creating the code. The data structure’s strengths and weaknesses are then assessed and distilled into a report. The researcher examines Melody’s code construction to discover what kind of information it is most effective at analyzing or categorizing, as well as whether the code tags and organizes information or clusters information into meaningful relationships.

 

This helps researchers understand what kind of questions these information structures should be applied to, and where a particular researcher’s methods might excel. The ability to use Quirkos to build and analyze unique and flexible databases from these structures has given Melody an edge in developing and sharing insights throughout the team.

 

While Melody Truckload’s app currently wraps up beta testing with commercial partners, the Quirkos approach has been put to the test most recently on the Melody team’s latest project, Melody Fashion.

 

“In the complex world of L.A.’s Fashion District, which is the part of town that houses the city’s fashion industry wholesale market, freight consolidation desperately needs to be modernized,” Kinzie said. “The objective of Melody Fashion is to provide a platform for fulfilment and consolidation that takes into account a detailed understanding of a market with many players.”

 

To that end, sales agents were trained to analyze interactions using grounded theory on Quirkos and to aggregate data garnered in their interactions with customers. It led to valuable insights, including a partnership with local shipping experts to bring Melody Fashion’s technology to the district.

 

Melody operations manager Marcus Galamay, who introduced new agents to Quirkos software and guided them through their first qualitative exercises, said, “Quirkos provides an intuitive introduction to qualitative analysis for our sales agents, augmenting their role in a way that’s expanded our insights into our client base. It’s a niche that many might not think to pursue, but it’s already delivered results in terms of better understanding of the data we generate and refining our market strategy based on that.”

 

Thanks to its ease of use and its powerful ability to assist in important social research, Quirkos was instrumental in providing Melody with the insight necessary to build smart and useful technology for a distinct and totally new customer base.

 

 

Using properties to describe your qualitative data sources

Properties and values editor in Quirkos

In Quirkos, the qualitative data you bring into the project is grouped as 'sources'. Each source might be something like an interview transcript, a news article, your own notes and memos, or even journal articles. Since it can be any source of text data, you can have a project that includes a large number of different types of source, which can be useful when putting your research together. This means that you can code things like your research questions, articles on theory, or even grey literature, and keep them in the same place as your research data.


The benefit of this approach is that you can quickly cross-reference your own research together with written articles, coding them on the same themes so you can compare them. However, there will be times that you only want to look at data from some of your sources. Perhaps you only want to look at journal articles written between a certain period, or look at respondent's data from just one city. By using the Source Properties in Quirkos, you can do all this and more: it allows you an essentially unlimited number of ways to describe the data. You can then use the query view to see results that match one or more properties, and even do comparisons. This Properties-Query combo is the best way to examine your coded qualitative data for trends and differences.

 

This article will outline a few different ways that you can use the source properties, and how to get the most use out of your research data and other sources.


When you bring a data source into Quirkos, the computer doesn't know anything about it. It's good practice to describe it, using what is sometimes called 'metadata' or 'data about data'. So for example, respondent data might have values for Age, Gender, Location, Occupation, Purchasing Habits... the list is endless. Research papers and textbooks will have values like Journal Name, Pulbication Year, Volume, Author, Page number etc.

 

Each of these categories in Quirkos are called 'Properties' and the possible data belonging to each property are called 'Values'. So for example, the Age of a respondent is a Property, and the value could be 42. Quirkos lets you have a practically unlimited number of Properties that describe all the sources in a project, and an unlimited number of Values.


The values can also be numerical (like age in years), discrete (like categories for Old, Young or 20-29) or even comments (like 'This person was uncomfortable revealing their age'). Properties can even have a mix of different data types as values.


To create properties and values in your project, click on the small 'grid' button on the top right corner of the screen. This toggles the properties view, and will show you the properties and values for the data source you are currently viewing. To look at a different source, just select it from the tabs at the bottom, or the complete list of sources in the source browser button (bottom left of the source column).


One here, you can create a new property and value with the (+) button at the bottom of the column, or use the 'Properties and Values Editor' to add lots of data at once, or to remove or edit existing values. The Editor also gives you the option of rearranging Properties and Values, and changing a Property to be 'multiple-choice' will let you assign more than one Value to each Property (for example to show that a person has multiple hobbies).


There are also a couple of features that help speed up data entry, for example the Properties Editor also allows you to create Properties that have pre-existing common values, for example 'Yes/No' properties, or common Likert Agree-Disagree scales. To define values for a property, use the orange drop-down arrow next to each Property. When you click on this, you can see all the values that have already been defined, as well as the option to add a new value directly.


I always try and encourage people to also use the properties creatively. You can use them to quickly create groups of your sources, and explore them together. So you may create a property for 'Unusual case', select Yes for those sources, and see what makes them special. There might even be something you didn't collect survey data for, but  is a clear category in the text, such as how someone voted. You can make this a Property too, and easily see who these people are and what they said. They can also be process-based properties: 'Ones I haven't coded Yet' or 'Ones I need to go over again'. Use the properties as a flexible way to manage and sort your data, in anyway you see fit! You can of course create and remove properties and values at any stage of your project, and don't forget to describe the 'type' of source: article, transcript, notes etc.


When you want to explore the data by property, use the Query view. This lets you set up very simple filters that will show you results of coded data that comes from particular sources. You can even run two queries at once, and see the results side-by-side to compare them. While by default the [ = ] option will return sources that match the value, you can also use 'Not equal' [!=] and ranges for numerical or alphabetic values ( < > etc). It's also possible to add many queries together with a simple interface, to create complex filters. So for example you can return results from just people between the ages of 30-35, who are Male, and live in France OR Germany.

 


This was a quick summary of how to describe your qualitative data in Quirkos: as always you can find more information in the video guides, and ask us a question in the forum.