How to Get Started with Amazon QuickSight

Amazon QuickSight is AWS's managed business intelligence service. You connect it to a data source, build a dataset, and generate charts and dashboards in a browser without installing anything. The Standard edition costs $9 per author per month and handles basic visualization. The Enterprise edition ($18/author/month) unlocks Amazon Q, which lets users type questions in plain English and get charts back instantly. This guide walks through the full setup from account creation to your first natural language query.

What You Need Before You Start

You need an active AWS account. If starting fresh, create one at aws.amazon.com and pick a region close to your users (us-east-1 is a safe default). QuickSight's free trial gives 30 days of Enterprise access, so you can test Amazon Q before committing to a paid plan.

IAM permissions matter. The AWS user or role signing up for QuickSight needs quicksight:* permissions. If you are using an existing AWS account where IAM is locked down, ask your AWS administrator to attach the AmazonQuickSightFullAccess managed policy to your IAM user before you begin.

Step 1 - Sign Up for QuickSight

Go to the AWS Console, search for "QuickSight," and click "Sign up for QuickSight." You will be prompted to choose between Standard and Enterprise. For teams that want Amazon Q natural language queries, select Enterprise. For personal testing on a small dataset, Standard is sufficient.

During signup, QuickSight asks for an account name, which is a unique identifier for your organization, and a notification email address. These cannot be changed after setup, so choose carefully. QuickSight also asks which AWS services it can access. At minimum, enable S3 access so you can upload your own files.

Step 2 - Connect a Data Source

QuickSight supports more than 20 data connectors. The most flexible starting path for AWS users is S3 combined with Athena, which lets you query flat files stored in a bucket using SQL without running a database server.

Connecting via S3 and Athena

Upload your CSV or Parquet files to an S3 bucket in the same region as your QuickSight account. Open the Athena console, create a database, and run a CREATE TABLE statement pointing to the S3 path where your files live. Athena uses AWS Glue Data Catalog under the hood, so your table schema becomes queryable immediately.

Return to QuickSight. On the Datasets page, click "New dataset," then select Athena as the data source type. Enter a name, select the Athena workgroup (the default workgroup works for most setups), and validate the connection. Choose the database and table you created in Athena. QuickSight will preview the first few rows. Select "Import to SPICE" for fast queries if your dataset is under 10 GB.

Connecting a CSV directly

If you have a file under 1 GB, you can skip Athena entirely. On the Datasets page, click "New dataset" and choose "Upload a file." QuickSight auto-detects column types. You can rename columns and change data types before saving the dataset.

Other supported connectors include RDS, Redshift, Aurora, MySQL, PostgreSQL, Snowflake, Databricks, and several SaaS sources like Salesforce. Once a dataset is ready, QuickSight shows a green status indicator along with row count and field count.

Step 3 - Build an Analysis

From the Datasets page, click your dataset and then "Create analysis." QuickSight opens the visual editor.

The left panel shows all fields in your dataset. Drag a dimension (a text or date field) to the X axis drop zone and a measure (a numeric field) to the Value drop zone. QuickSight picks a chart type automatically; you can override it using the visual type panel on the left.

A few defaults to configure immediately: change the date aggregation to match your reporting cadence (daily, weekly, or monthly) using the date field dropdown. Add filters by clicking the Filter pane and selecting "Add filter." Filters can apply to the entire sheet or just the selected visual. The Insight panel (the lightbulb icon) auto-generates a written summary of what each chart shows, and that summary updates as filters change.

Step 4 - Enable Amazon Q for Natural Language Queries

Amazon Q in QuickSight lets non-technical users type questions like "What were sales by region last quarter?" and receive an auto-generated chart as the answer. This requires Enterprise edition and a configured Q topic.

Creating a Q topic

From the QuickSight home page, click "Topics" in the left menu, then "New topic." Give it a business-friendly name such as "Sales Performance" and select the dataset or datasets it should cover.

QuickSight then prompts you to add synonyms and friendly names to your columns. This step is worth the time. If your column is named rev_ttl_usd, add the synonym "revenue" so users do not need to know internal field names. Add three to five sample questions and mark the expected answers as correct. This teaches Q how your data is structured and significantly improves accuracy.

Using the Q search bar

Add the Q search bar to any dashboard by opening the analysis, clicking "Add" in the top toolbar, and selecting "Q search bar." Resize and position it at the top of the sheet. Publish the dashboard, and users with QuickSight reader accounts can type questions directly into the bar.

Amazon Q generates a chart, a written insight, and a confidence indicator. If confidence is low, the system flags that the question may be ambiguous and offers alternative interpretations. In practice, Amazon Q handles aggregation questions well: totals, averages, and trends over time across one or two dimensions. Multi-hop questions that require joining logic across several tables still require a data author to pre-configure the topic relationships manually.

According to AWS, teams using Amazon Q topics with at least 10 configured synonyms see answer accuracy rates above 80 percent on common business queries, versus roughly 60 percent with no synonym configuration.

Step 5 - Share Dashboards with Your Team

Publish the analysis as a dashboard by clicking "Share" in the top right and then "Publish dashboard." Anyone with a QuickSight reader account ($4/user/month on Enterprise) can view published dashboards. Readers cannot edit visuals but can apply filters, export data, and use the Q search bar.

QuickSight also supports row-level security, which restricts each user to the rows in the dataset they are allowed to see. Configure RLS by uploading a permissions table that maps user email addresses to filter values. For example, a sales representative can be restricted to seeing only their own region's data within a shared dashboard.

What Most Guides Leave Out

The biggest friction in new QuickSight setups is the SPICE refresh schedule. SPICE is an in-memory store, not a live connection. If you import data into SPICE, the dashboard shows data as of the last refresh. Set a refresh schedule on the dataset immediately after importing. Daily refreshes are available on Standard; hourly refreshes require Enterprise.

QuickSight is also a regional service. A dataset created in us-east-1 is not visible in ap-southeast-1. If your team is distributed across AWS regions, centralize QuickSight in one region and grant cross-region access to your S3 buckets, or manage separate QuickSight accounts per region.

If your data is not in AWS yet and you want analysis without any infrastructure setup, VSLZ lets you upload a CSV, ask questions in plain English, and get charts and statistical analysis back without configuring cloud resources.

Practical Summary

QuickSight takes 20 to 30 minutes to reach a working dashboard with a real dataset. The S3 plus Athena path is the most flexible starting point for AWS users. Amazon Q natural language queries require Enterprise edition and a Q topic with configured field synonyms. Set a SPICE refresh schedule immediately after importing any dataset to avoid serving stale numbers to your team.