Grow offers a new way to pull data in and build metrics. This is done through creating datasets. Datasets are reusable report configurations. Using a dataset allows you many benefits, including not having to reconfigure the same report over and over for multiple metrics. This article will give you an overview of this feature and what you can do with it.
What is dataset?
A dataset is exactly what it sounds like---a set of your data. When you create a dataset, you pull in your data and save it as an existing configuration. Grow continually updates that data, and you can access the dataset at any time and build metrics from it.
What can I do with datasets?
- Power your metrics from the same set of data. You just have to set up the dataset once, then Grow does the rest to keep that data up-to-date. This means that you can build metrics more quickly because you don’t have to configure as many reports, and you don’t have to wait for slower integrations to return your data each time.
- Analyze larger quantities of data. Datasets and Grow’s new metric builder make it easier to store and analyze larger sets of data. You can dig deeper into your data and get more meaningful and actionable insights out.
- Pull the freshest data into your metric. Datasets are set up to make API calls throughout the day, and offer you the most up-to-date information.
How do I store a dataset?
This article will walk you through creating a dataset in Grow.
Can I view my dataset after I’ve created it? What about exporting the data?
You can view your dataset after you’ve created it. To do this, go to the list of datasets you have access to. (Click Add Metric and select New Metric Builder.)
When you hover over the dataset, you will see two actions appear at the right-hand side: a blue Create Metric button, and an eye. Clicking on the eye will pull up a preview of the data.
Currently, exporting the data from a dataset is not supported. You can, however, create a metric with your dataset, then export the data from the metric on your dashboard.
Where can I find the datasets I have access to?
On the datasets page. To get to this list, log in to your Grow account and click on Add Metric. (Select Use New Builder if you are given options.) This will bring you to the list of datasets that you have access to. You can build metrics or create a new dataset from this page.
How often does my data refresh?
By default, datasets are set to refresh every hour. Once that data updates in the dataset, your metrics will reflect the change within 5 minutes of the dataset being refreshed.
If you want to see the last time your data was updated, go to your list of datasets. There is a column titled “Last Updated” and each dataset shows the last time they updated.
How do dataset refreshes work?
There are two ways that Grow can update your dataset: ripping and replacing, and appending. The update strategy we use depends on how data comes back from each data source.
For example, if all of the data is returned with one API call, we use a method called “Rip and Replace.” A common use case for this is any dataset that is based on a spreadsheet. Each time we update that dataset, we just replace all of the data we previously had with the current data from that source.
The second scenario is when we are able to split up the data into a series of calls that use date parameters in the request. In this case, we make incremental API calls for smaller chunks of dates, and append the new entries onto the dataset as it gets built out. By doing this, we are able to maintain the original data by only appending the recent changes to that set of data, rather than replacing the whole dataset each time it needs to be updated.
What if the schema of my data changes? Will it show up in my dataset?
If columns are added or removed from your dataset after the initial creation, rip and replace datasets will reflect the new set of columns. Append datasets will ignore any of these changes, and only display the rows and columns that were available when the dataset was initially created.
Questions about datasets? Email us at email@example.com or chat in on the site.