In partnership with

⚒️ Cool Tool - Deel

⚒️ Cool Tool (2) - 1440 Media

🔎 Deep Dive: ODAG (On-Demand App Generation) in Qlik Sense - Cloud vs. On-Prem

Struggling to explore massive datasets in Qlik without overwhelming memory limits? That’s where ODAG comes in.

ODAG (On-Demand App Generation) is Qlik’s built-in mechanism for working with large-scale data without compromising performance.

It helps users drill into huge datasets - like multi-terabyte transaction logs or high-volume IoT records - by generating apps on the fly with just the data they need.

This week, we’re breaking down how ODAG works in both Qlik Sense Cloud and Qlik Sense Enterprise (On-Prem).

You’ll learn where it fits best, how to set it up, what’s different between cloud and on-prem, and how to put it to work in your enterprise environment.

🔍 Why This Matters: App Generation at Enterprise Scale

ODAG was built for one specific job: reducing data volume by segmenting analysis workflows.

Instead of trying to load everything into memory, ODAG allows users to select a subset of data (e.g., a region, date range, or customer segment) and generate a new Qlik app on demand that contains just that slice of data.

This keeps memory usage lean, performance high, and the user experience focused.

Common ODAG use cases:

• High-volume sales or transaction detail reporting

• Drill-downs from summary dashboards into record-level views

• Exploring longitudinal data (e.g., patient journeys, customer behavior)

• Reducing memory usage when working with wide or deep fact tables

• Supporting data lakes or warehouse offloading scenarios

🛠️ How ODAG Works (At a High Level)

1. The Selection App: This app presents users with high-level filters and KPIs (e.g., total sales by region and product).

2. The Navigation Link: Users make selections and trigger an ODAG link, which passes those parameters to a template app.

3. The Template App: This app contains a load script with bind variables. When ODAG is triggered, Qlik generates a new app using the selections and the script logic.

4. The Generated App: A brand-new Qlik app is created with just the relevant data subset. Users can explore it freely, share it, or schedule it.

   

⚖️ Cloud vs. On-Prem: Key Differences

While ODAG core concepts are consistent, the setup and experience diverge significantly between Qlik Sense Cloud and On-Prem deployments.

🔧 ODAG Setup: Step-by-Step (Cloud)

1. Build the Selection App: Load the summary data and build filters.

2. Create Template App: Define the full data load with binding variables (e.g., SET vRegion = '$(odag_Region)').

3. Create ODAG Link:

   1. Open the selection app in Qlik Cloud.

   2. Click the "..." menu and select “ODAG links.”

   3. Click "Create new ODAG link," choose your template app, and define field mappings (e.g., odag_Region).

   4. Optionally configure naming, tags, and target space.

   5. Save the link to activate navigation.

4. Test Navigation: Make a selection and launch the link - a new app should be generated based on your criteria.

   

5. Automate (Optional): Use Qlik Automate to reload the app or notify users.

🔧 ODAG Setup: Step-by-Step (On-Prem)

1. Create Selection App: Includes basic filters and KPIs.

2. Configure QMC and ODAG Link:

   1. A. In the QMC, go to ODAG settings and define a new navigation link.

   2. B. Choose the selection app and target template app.

   3. C. Map selection fields to bind variables (e.g., odag_DateFrom).

3. Create Template App: Include variables in the load script (e.g., LET vCountry = '$(odag_Country)').

4. Publish and Test: Publish both apps to streams, make selections, and verify that the ODAG link correctly generates a filtered child app.

⚠️ Limitations & Considerations

• ODAG does not support associative exploration across datasets - it creates a standalone app.

• Template apps must be designed carefully with parameterization in mind.

• Qlik Cloud simplifies much of the process, but doesn’t yet expose all the advanced controls available on-prem.

• Consider storage and governance: ODAG can generate a large number of apps over time if not managed.

• ODAG requires stable bind variables and must avoid script loops and circular logic.

🏬 Enterprise Use Case: Regional Drilldowns in Retail

A global retailer uses ODAG to allow regional managers to drill into transactional data. The selection app contains sales KPIs by region and week. Managers can select their region and generate a detailed app that loads SKU-level sales for just their territory. This avoids loading 100M+ rows across all regions every time.

• Selection app size: < 200 MB

• Generated app size: 300 MB to 2 GB depending on region

• Speed: Apps generated in < 30 seconds with filters applied

💡 Top Tips for ODAG Success (Cloud & On-Prem)

For Qlik Cloud:

• Use automation to clean up old generated apps to avoid clutter and storage issues.

• Use ODAG tags and metadata to organize generated apps by user or purpose.

• Test bind variable logic with static inputs before deploying in production.

For Qlik On-Prem:

• Build PowerShell scripts or use the QRS API to monitor generated apps and clean up periodically.

• Consider dedicated streams or app folders to manage ODAG outputs.

• Implement ODAG testing as part of your deployment QA - especially where reload triggers are external.

General ODAG Advice:

• Keep selection apps lightweight and visually simple.

• Limit the number of ODAG fields passed in a single link to reduce script complexity.

• Always validate selections with preview KPIs before triggering navigation.

🧠 Final Thoughts

ODAG remains one of the most underused yet powerful features in Qlik’s stack.

In both Cloud and On-Prem environments, it empowers business users to explore massive datasets without the technical overhead of full-scale app design or constant IT intervention.

If you're designing enterprise Qlik solutions, especially those involving heavy datasets, ODAG should absolutely be part of your toolbox.