/automation_for_ur_at_scale

[Presentation] Code presented during my talk to the Spotify Product Insights team (February 2020)

Primary LanguageJupyter Notebook

Automation for UR at Scale

This is code that I presented during my talk for the Product Insights team at Spotify (February 2020).

Right-click here to view the slide deck in a new tab.

I have included the survey and scripts used in the example so that anyone interested could try it themselves. The scripts interact with the Qualtrics API to: one, automatically embed user data into a Qualtrics survey, and two, automatically create clean and interpretable data exports.

To try the example:

  1. Download survey.qsf and import it into Qualtrics

    • Once imported, make sure you hit the publish button.
    • Notice that all of the questions and response choices are labeled - this helps make the data exports much easier to read!
  2. Open the Github repo. You can do this in two ways:

    • If you have Jupyter Notebooks on your computer: clone the repo and navigate to the scripts folder in Jupyter.
    • If you do not have Jupyter Notebooks: click here → Binder to launch the code using a website called Binder. Binder will take a few minutes to load, but once it loads, you'll have everything you need to run the scripts!
  3. Navigate to the scripts folder and open embed_user_data_into_survey.ipynb. Once you have the code open:

    • Enter your Qualtrics IDs.
    • Ensure that you have access to the Qualtrics API.
      • If you do not, you'll see <Response 403> when you try to embed the variable headers and it will not work.
      • I had to contact my IT administrator in order to receive access.
  4. Once you have created your embedded data fields using the code above, create some fake data!

    • Go to the data folder and open survey_distribution_list.xlsx. If you opened the code using Binder, download survey_distribution_list.xlsx and open it on your computer.
    • Click on a few of the links and fill out the survey to generate your fake data.
  5. Navigate back to the scripts folder and open export_survey_data.ipynb. Before running through the code, make sure to enter your Qualtrics IDs.

Helpful links from Qualtrics: