# Create a dataset Datasets provide a structured way to store and manage data, ensuring that only the most relevant and valuable information is included for reporting. By curating datasets to focus on key metrics and dimensions, you can gain deeper insights, track performance more effectively, and customize visualizations to align with your specific business needs. Whether you're working with manually uploaded data or integrating with external sources, datasets give you greater flexibility and control over how your data is used in Databox. ## Create a dataset You can create a dataset from the following locations: - [Data Manager > Datasets](https://app.databox.com/data-manager/datasets) → Click **+ New Dataset** - [Account Management > Data Manager > Datasets](https://account.databox.com/data-manager/datasets) → Click **+ Add Dataset** After selecting an existing data source or adding a new one, the dataset setup window will open. In this window, you'll first need to select the data to include in the dataset, and then specify its data update frequency. ### Select data In this stage, you will choose the view and associated columns to be included in your dataset. - **Views**: These are categories of data that typically correspond to specific objects, modules, or components within your data source. Examples include: - Deals (HubSpot) - Orders (Shopify) - Subscriptions (Stripe) - Invoices (QuickBooks) - **Columns**: These refer to the individual fields or attributes within each view. For instance: - Deal Amount (HubSpot) - Order Date (Shopify) - Customer Email (Stripe) - Invoice Balance (QuickBooks) If you are working with a compatible SQL integration, you can also enter a custom query, and its output will be used to create the dataset. This approach is advisable when more complex transformations are necessary to generate the desired dataset. ### Update frequency Next, you'll select how often Databox should refresh the data in the dataset. The available synchronization frequencies depend on the technical capabilities of the chosen integration and [your account subscription](https://app.databox.com/settings/billing). update-frequency Once you've chosen the desired frequency, click **Save** to finalize the setup and initiate a full synchronization of the selected view and columns in the backend. Databox will generate a preview as soon as the data becomes available. If syncing takes longer than 30 seconds, a message will appear, and you'll receive an email notification when your data is ready. ## Curate the dataset To ensure your dataset includes all relevant data for reporting: - **Define key metrics and insights** the dataset should support. - **Identify critical data points** needed to calculate those metrics, including: - **Relevant measures** (e.g., sales, revenue, conversions). - **Dates** for tracking trends over time. - **Dimensions** for segmenting data (e.g., region, product category). - **Additional data** necessary for drill-down or deeper analysis. To create metrics from a dataset, at least one **datetime** column must be present, as it serves as the basis for the date range selector within the application. It's equally important to ensure the dataset captures granular data where needed while excluding unnecessary fields that could add complexity without contributing value. By focusing on the right data, you can ensure the dataset remains both comprehensive and efficient for reporting purposes. ## Edit the dataset To modify your dataset, click **Edit Data** in the view mode. Then, click the **down arrow** (![chevron-down](/assets/chevron-down.dfd0e90adc6eb576be32262463090aae7c7f5408ed7975e3795f69775e36d174.419bb737.svg)) next to a column name to access the actions menu. The following options are available: - **Set data type:** Choose a new type for the column if there is a mismatch or conversion is needed. - **Filter:** Add one or more filters to include or exclude specific rows. - **Insert column left:** Add a new calculated column to the left of the selected column. - **Insert column right:** Add a new calculated column to the right of the selected column. - **Hide:** Remove the column from view within the app without deleting it. You can also **rename columns** by clicking the column name box and entering a new name. Once you're satisfied with your changes, click **Save** in the top-right corner of the screen. Yes, it's possible to combine multiple datasets into one, provided they have at least one column with matching values. For more details, refer to [this article](/merge-datasets). No, only one filter type (either AND or OR) can be applied to a column at once. Each dataset can hold a maximum of 100 columns, and its total size cannot exceed 200MB.   Ask Genie Get instant answers or help with your data using the in-app AI assistant. Talk to an expert For customers: Get help with your setup, strategy, or making the most of Databox. Book a demo New to Databox? See how it works and get guidance on getting started. Send an email Reach out to support for help with your account, data, or technical issues.