The primary cause of slow performance is the execution of row-level string functions (Left, Trim, Right) and arithmetic operations within a front-end chart expression on a very large dataset (10 million rows). These calculations are performed every time the app state changes. The most effective Qlik performance optimization strategy is to move these transformations and calculations from the chart expression into the data load script. By pre-calculating the required values during the data load, the results are stored directly in the data model. The front-end KPI then only needs to perform a simple aggregation on this pre-calculated field, which is significantly faster and less resource-intensive. Option B correctly describes this best practice using the Data Manager.

A. This proposes an external solution outside the Qlik business analyst's control and does not use Qlik Sense capabilities to resolve the immediate performance issue within the app.

C. This keeps the calculation in the front-end master measure. While SubField might be slightly more efficient than nested functions, it does not solve the core problem of performing row-level calculations at runtime.

D. This action only modifies the string and does not create the final calculated numeric field. The front-end expression would still need to parse the modified string, failing to offload the calculation to the script.

---

1. Qlik Sense® on Windows Help, "Tips for improving app performance", section "Move calculations to the data load script". This official documentation explicitly states: "Calculations that you make in chart expressions are made every time you make a selection in the app. This can affect the response times in your app. To improve performance, you should move calculations to the data load script when possible." This directly supports the methodology described in the correct answer.

2. Qlik Sense® on Windows Help, "Managing data with Data manager", section "Adding a calculated field". This document outlines the procedure for creating new fields based on calculations during the data preparation stage, which is a core part of the solution in option B. It confirms that a business analyst can perform this task using the Data Manager interface.

3. Qlik Sense® on Windows Help, "Managing data with Data manager", section "Splitting a table field". This source details how to use the Split transformation in the Data Manager to parse a single field into multiple fields based on a delimiter, which is the first step described in the correct answer.