2 March 2023 Parsing JSON data in Qlik Sense Share this message After a little break last January, we’re back with our weekly Friday Qlik Test Prep question (follow us on LinkedIn if you want to participate). This time, we have a question about parsing JSON data in Qlik Sense. We’ll first look at the question and the correct answer. Next, we’ll take a look at some of the basics to consider when working with JSON data in Qlik Sense. While most of the answers might look (somewhat) plausible there was only one correct answer: The correct answer is B: JsonGet Technically, this answer isn’t entirely correct as this function returns "Bitmetric" instead of Bitmetric. The double quotes that encapsulate the string are included. One way to clean this up is to use the PurgeChar() function: PurgeChar(JsonGet(OrderDetails, '/shipTo/name'), '"') Let’s have a closer look at how the JsonGet() function works. The Qlik JsonGet function The JsonGet() function can be used in both charts and scripts. It takes two arguments: A string containing JSON data; The ‘path’ to retrieve from the data. This can be a single value, but also a subset of the JSON data. We’ll explain this in the next section. The function returns the path. If invalid JSON was provided, or if the requested path does not exist, the function returns null. The JSON path, or pointer The JSON path, or pointer, specifies the ‘route’ to the data that you want to retrieve. Consider the following JSON data for a book and its buyer: If we want to retrieve the name of the country where the book was shipped, we start at the root of the data, then go to the shipTo key and finally to the country key: Each ‘hop’ is separated with a forward slash (/) and we always start the path with a slash to indicate that we’re starting at the root. In our example, it means that we use the following path: /shipTo/country Referencing JSON data in arrays with an index Not all data in JSON can be directly referenced by a key. In our example, the authors key contains an array with multiple authors. We can retrieve the names of the individual authors by referencing the index of the array. The index of the array starts at 0. To get the name of the first author, Angelika Klidas, we use the following path: /authors/0/name Note that we used 0 to reference the index. A few more examples Based on the same example data, here are some more paths and their result: PathResultResult type/title“Data Literacy in Practice”string/price29.99number/authors/1{“name”: “Kevin Hanegan”}object/shipTo/name“Bitmetric”string/shipTo/shippedtrueboolean Other useful JSON functions in Qlik There are two additional JSON functions in Qlik: IsJson() tests whether the supplied string contains valid JSON data. Optionally, it can also be used to test for specific data types. JsonSet() allows us to update JSON data in a string, or to add new values to it. Since the main focus of this article is extracting data from JSON, we’ll only expand on the first function for now. The Qlik IsJson function The IsJson() function tests if a string contains valid JSON data. This is very convenient for error handling and generally checking our data before we try to parse it. Consider the same example data as before, stored in a string called OrderData: Using the following expression we can test if the JSON data is valid (it is). IsJson(OrderData) This only tells us that it is valid JSON though, and not what specific data type it is. To test for specific data types, we use the second argument of the IsJson() function. Testing for specific JSON data types We can test for the following JSON data types: ArgumentDescriptionExample where truevalueAny valid JSON. If no argument is supplied this is used by default.IsJson(OrderData)objectAn object, contained between { and }IsJson(JsonGet(OrderData, '/shipTo'), 'object')arrayAn array, contained between [ and ]IsJson(JsonGet(OrderData, '/authors'), 'array')stringA string valueIsJson(JsonGet(OrderData, '/title'), 'string')numberA numeric valueIsJson(JsonGet(OrderData, '/price'), 'number')booleanA boolean valueIsJson(JsonGet(OrderData, '/shipTo/shipped'), 'boolean')nullA null value. Note that in JSON null values must be explicitly represented using the null valueIsJson(JsonGet('{"key":null}', '/key'), 'null') And that’s it for the basics of parsing JSON data in Qlik Sense. So what’s next? Over the past decade, JSON data has become more and more common. At Bitmetric, almost all our clients have at least one source that provides data in this form. Two use cases of these techniques that we’ll be covering in more detail later on this blog are: Storing configuration settings in JSON Extracting data from REST APIs, and avoiding a lot of the hassles with the Qlik REST Connector generating dozens of tables. That’s it for this week. See you next Friday? More from the Bitmetric team Bitmetric Qlik Support Keep your Qlik environment running smoothly with proactive support that prevents issues before they appear. Available on flexible monthly plans. Learn more. Qlik vs Power BI Series See how Qlik and Power BI perform when theory meets reality. Three apps rebuilt, every step examined. Learn more. API Friday Qlik Test Prep Qlik Script How can we help? Whether something’s still unclear or you’re ready to take the next step, Barry and Eric are happy to talk it through. Email us, call us, or book a meeting at a time that works for you. Call us Mail us 1 September 2026 Qlik Answers review: What we learned from real-world testing We tested Qlik Answers on a real production sales model to see how it handles real-world analytics. The results were promising, but getting reliable answers required careful work on the logical model, master measures, dimensions and business context. AI Data Analysis Qlik Semantic Layer 25 August 2026 Your semantic layer is becoming the API for AI Semantic layers are not new, but AI is giving them a new role. Where they once primarily powered dashboards, semantic models are increasingly becoming the layer that provides AI with reliable business context. AI Data Governance Microsoft Fabric Qlik Semantic Layer TimeXtender 17 August 2026 Putting AI to work on your data works. But not like this. Putting AI to work on your data sounds simple. But what can we learn from organizations already doing it? Based on Anthropic’s experience and independent research: what works, what doesn’t, and why maintenance is where the real work lies. AI Data Governance Data Management Power BI Qlik
1 September 2026 Qlik Answers review: What we learned from real-world testing We tested Qlik Answers on a real production sales model to see how it handles real-world analytics. The results were promising, but getting reliable answers required careful work on the logical model, master measures, dimensions and business context. AI Data Analysis Qlik Semantic Layer
25 August 2026 Your semantic layer is becoming the API for AI Semantic layers are not new, but AI is giving them a new role. Where they once primarily powered dashboards, semantic models are increasingly becoming the layer that provides AI with reliable business context. AI Data Governance Microsoft Fabric Qlik Semantic Layer TimeXtender
17 August 2026 Putting AI to work on your data works. But not like this. Putting AI to work on your data sounds simple. But what can we learn from organizations already doing it? Based on Anthropic’s experience and independent research: what works, what doesn’t, and why maintenance is where the real work lies. AI Data Governance Data Management Power BI Qlik