Amazon QuickSight で高チャートを使い多地域可視化

When your carrier performance data spans multiple regions, your dashboard must reconcile fundamentally different competitive structures within a single view. For example, in the US, you rank three carriers (Carrier 1–3) across 49 states and hundreds of metro markets. In the UK, you’re comparing four carriers (Carrier 4–7) across a separate set of national regions. The challenge you face is not only geographic but also structural.

The differences in carrier count, market density, and regional detail between these two geographies mean that no single out-of-the-box chart type can tell your full story. You need visualizations that answer layered questions simultaneously: who leads, by how much, how consistently, and whether they are improving or declining.

You will find that native Amazon Quick visualizations do not tell your full story when you try to answer these questions at scale. Amazon Quick Sight can’t encode which carrier or combination of carriers dominates each geographic tile. It can’t render a radar chart that profiles all seven carriers across seven performance categories simultaneously. And it can’t produce chart types such as bullet charts and dumbbell charts, which you need to surface the nuances in your data.

Every time you try to approximate these visuals in native Quick Sight, you end up building separate dashboards per Region, using stacked bar workarounds for deltas, or averaging metrics that mask volatility between a carrier’s best and worst markets. These compromises add maintenance overhead and obscure the insights your stakeholders need to make confident, data-driven decisions.

By embedding Highcharts as a custom visual in Quick Sight, you address these limitations with a single, unified JSON configuration. With its tilemap chart type and colorAxis.dataClasses, you can encode carrier dominance (including multi-carrier ties) across multiple regions in one visual, with legend toggles for region filtering. Its polar line chart renders a true polygon radar for multi-carrier, multi-category comparison, giving you the full competitive profile.

Beyond geographic and radar views, Highcharts unlocks the chart types your analysis demands. You can use bullet charts to show performance against a 100-point target with color-banded zones, dumbbell charts to expose the spread between a carrier’s best and worst metro market, and waterfall charts to break down period-over-period score deltas.

This post shows you how to build multi-Region carrier performance dashboards in Quick Sight using Highcharts custom visualizations to overcome native chart limitations. You will learn how to maintain data sovereignty across AWS Regions while creating unified visualizations through the Quick Sight federated dataset capability. The solution includes production-ready chart configurations and addresses security, compliance, and scalability requirements.

Solution overview

Your architecture depends on factors such as data residency requirements, latency needs, and compliance obligations. Two architectural patterns address this:

Option 1: Single AWS Region

With this approach, you store carrier data (for example, both US and UK) in one AWS Region (for example, us-east-1 or eu-west-1), using a single Quick Sight SPICE (Super-fast, Parallel, In-memory Calculation Engine) dataset with a country column to logically identify the data.

While this architecture is simpler for you to manage with one dataset and one SPICE refresh, it carries a potential drawback: if you choose a US-based Region, your data physically resides outside the UK, which may conflict with the UK General Data Protection Regulation (GDPR) or data residency rules if your dataset contains personally identifiable information (PII).

Option 2: Separate AWS Regions per country

Sometimes, data sovereignty laws require you to physically store and process certain categories of data (particularly data about individuals, network usage patterns, and commercially sensitive performance metrics) within the legal jurisdiction where you collected the data. For example, for UK carrier data, the UK GDPR (retained post-Brexit) mandates that personal and operationally sensitive data about UK consumers and UK network performance must reside on infrastructure located within the UK or the broader European Economic Area. This means that you must place US carrier data in a US AWS Region (for example, us-east-1 or us-west-2) and UK carrier data in an EU-adherent Region, with no co-mingling of raw data across those boundaries. The two datasets are sovereign. You can’t merge them into a single Amazon Simple Storage Service (Amazon S3) bucket without violating one or both jurisdictions’ requirements.

Despite the data residing in separate AWS Regions, you can build a single unified Highcharts dashboard using Quick architecture without moving raw data across borders. You use federated dataset joins at the Quick layer rather than at the storage layer: you ingest the aggregated US dataset (in us-east-1) and the aggregated UK dataset (in eu-west-2) into their respective regional SPICE instances, and the Data Prep capability of Amazon Quick appends them at data prep time into a single logical dataset. The resulting unified dataset feeds a single set of Highcharts configurations (tilemap, radar, bullet, dumbbell, waterfall, streamgraph) that render multi-Region carriers together. No raw sovereign data leaves its home Region. You federate only the aggregated RootScore, Rank, and ColorValue metrics computed within each Region into the unified view, satisfying data residency requirements while still enabling the combined visualizations your dashboard requires.

While Option 1 suits most scenarios, the post walks you through Option 2 because it addresses the more complex requirements that telecommunications companies frequently encounter.

Solution architecture

The following diagram shows how the multi-Region architecture federates carrier performance data while maintaining data sovereignty. Two regional data pipelines, one in us-east-1 for US carrier data and one in eu-west-2 for UK carrier data independently aggregate performance metrics within their respective Regions. At the Quick layer, the Data Prep capability federates these aggregated datasets at query time into a single logical dataset. This unified dataset feeds the Highcharts custom visuals, enabling a combined multi-Region dashboard without moving raw data across borders

Prerequisites

Before you begin, make sure that you have the following:

• An AWS account with access to Quick.

• Data sources configured across multiple AWS Regions.

• Basic familiarity with Quick Sight dashboards and data integration.

• Understanding of JSON configuration for Highcharts.

• AWS Identity and Access Management (IAM) permissions to access data sources across regions.

Data preparation

To aggregate data from multiple regions, configure your data sources in Quick. This step helps you establish the connections that enable cross-Region data retrieval.

• In the Quick console, navigate to Datasets from the left navigation panel.

• Choose Create data source to begin adding your first regional data source.

• Select your data source type (such as Amazon S3, Amazon Relational Database Service (Amazon RDS), or another data source of your choice).

• Enter the connection details for your first region, including endpoint URLs and authentication credentials.

• Choose Validate connection to verify data accessibility.

• After validating the connection, choose Create data source.

• Repeat steps 2–6 for each additional AWS Region that you want to include in your analysis.

After configuring your regional data sources, you can proceed to combine them into a unified dataset.

Configure data aggregation and transformation

After connecting your data sources, set up aggregation rules to combine data from multiple regions. This step helps maintain consistent data formatting and produces a unified analysis.

• Choose Datasets under Quick.

• Choose Create dataset.

• Select one of the data sources you registered in the previous section and its associated schema.

• Choose Edit/Preview data.

• In the Data Preparation interface, choose Add data.

• Select your second regional data source from the list.

• Choose Append to stack the datasets vertically.

• Verify that column names match between datasets (rename if needed).

• Add a calculated field named Region to identify data origin (for example, IF(condition, 'US', 'UK')).

• Preview the combined dataset to verify row counts.

• Add calculated fields to normalize data formats across regions if needed. For example, convert timestamps to a common time zone.

• Choose Save & publish to publish your aggregated dataset.

• After you publish the dataset, configure the refresh schedule for this dataset under Schedule refresh to keep your aggregated data current.

Set the frequency based on your business requirements (hourly, daily, or weekly).

You can now create Highcharts visualizations using your aggregate dataset.

To learn how to set up cross-region analytics data for your Highcharts visualizations, see this walkthrough on building a multi-Region analytics solution.

Dataset overview

The carrier performance dataset used in this walkthrough contains records spanning seven carriers across two countries (UK, USA). The dataset includes seven performance categories (Call, Data, Overall, Reliability, Responsiveness, Text, Video), with rankings and RootScore metrics across two measurement periods (2024-2H, 2025-1H).

The dataset schema consists of the following attributes:

Field

Description

carrier

Carrier ID (for example, Carrier1 through Carrier7)

category

Performance category (Call, Data, Overall, Reliability, Responsiveness, Text, Video)

country

Country of measurement (USA, UK)

rank

Competitive ranking (1–4) within that market and category

rootscore

Performance score (0–100 scale)

product_period

Measurement period (2024-2H, 2025-1H)

The following table shows a representative sample from the dataset:

carrier

category

country

rank

rootscore

product_period

Carrier1

Call

USA

1.0

99.53

2024-2H

Carrier1

Call

USA

1.0

98.08

2024-2H

Carrier1

Call

USA

1.0

82.18

2024-2H

Carrier1

Call

USA

1.0

99.99

2024-2H

Carrier1

Call

USA

1.0

98.79

2024-2H

Carrier1

Call

USA

1.0

98.37

2024-2H

Carrier1

Call

USA

1.0

99.96

2024-2H

Carrier1

Call

USA

1.0

99.12

2024-2H

Understanding dynamic field well bindings

The code blocks in this post show how data flows dynamically from the dataset by using placeholder tokens ({{token}}) that map to specific dataset queries.

Each code block includes two components:

• The field well mapping specifies which dataset fields (carrier, category, rootscore, rank, product_period, country) map to which field wells. These define how data flows from the dataset into the chart.

• The JSON template uses placeholder tokens (for example, {{carrier_list}} and {{avg_rootscore_period_1}}) that the Quick Sight field well engine resolves at render time. The mapping legend that follows each chart documents what each placeholder resolves to.

Highcharts implementation walkthrough

In this section, you walk through Highcharts visualizations that address the multi-Region carrier analysis requirements that native Quick Sight visuals can’t fulfill on their own. Before diving into the code, review the following requirements table. It maps each visualization to its Highcharts chart type, summarizes what the visual must accomplish, and explains why you cannot achieve the same result with built-in Quick Sight chart types. Each chart that follows includes a complete JSON configuration you can copy directly into your Quick Sight Highcharts custom visual, along with a rendered preview showing the expected output. Because every configuration binds dynamically to the Quick Sight field wells rather than hardcoding carrier names or region values, you can adapt these examples to your own dataset without rewriting the JSON. Only your lookup table needs to reflect your carriers and regions.

Requirements

#

Visual name

Highcharts type

Requirement summary

Why Highcharts

1

US Carrier Performance Radar

polar + line

Spider/polygon chart showing each US carrier’s average score across seven categories (Call, Data, Overall, Reliability, Responsiveness, Text, Video).

No polar or radar chart type in Quick Sight. You can’t create a multi-axis, multi-carrier overlay with a polygon grid natively.

2

Overlapping Column Chart with Target Bubbles

column

Compare 2024-2H and 2025-1H RootScore for all seven carriers in a single view. The 2024-2H column renders a wider solid bar. The 2025-1H column overlaps it as a semitransparent overlay. Orange bubbles at y=100 mark the target, and a dashed red reference line reinforces the benchmark.

No similar chart type in Quick Sight. If you use stacked bar workarounds, you lose the delta narrative entirely.

3

Variwide Chart: Win Share × Volume

variwide

Carrier1 win % as bar height. Total #1 wins in that category as bar width, thus two key performance indicators in one bar.

All Quick Sight bars have equal fixed width. You can’t encode a second dimension (market volume) into bar width.

4

Streamgraph: Win Flow

streamgraph

Flowing stream width per carrier showing #1 Overall win count across periods and reveals competitive momentum.

No streamgraph type in Quick Sight. The organic flowing proportional share shape is exclusive to Highcharts.

5

Hexagonal Tilemap for Market Share

tilemap

A hexagonal grid where each tile represents ~1% of markets won. Carriers fill the grid proportionally to their share, creating an immediately intuitive visual of competitive concentration.

Quick Sight built-in map visuals (filled maps, point maps) render geographic shapes or coordinates, not abstract hex grids. There’s no native “tilemap” or hex-bin chart type.

6

Combined US + UK Packed Bubble

packedbubble

All carriers in a single packed bubble configuration with distinct cluster regions per carrier.

No packed bubble type is available. You can’t group carriers into distinct cluster regions within one visual.

Create a Highcharts custom visual

To build each visualization, follow these steps:

• Navigate to Analyses and open your analysis.

• Add a Highcharts visual.

• Map the analysis to your prepared dataset.

• Update the JSON configuration in the format panel.

• Map your dataset fields to the field wells referenced in the JSON.

• Verify that the chart renders correctly.

Chart 1: Average overall RootScore, 2024-2H compared to 2025-1H

The following chart compares the average overall RootScore for all seven carriers across two measurement periods (2024-2H and 2025-1H). Each carrier displays two overlapping columns, a wider solid bar for the earlier period and a semitransparent overlay for the later period so you can immediately see period-over-period improvement or decline. Orange bubbles positioned at 100 mark the target score, and a dashed red reference line reinforces this benchmark. You can observe that Carriers 1–3 consistently score above 95 and approach the 10