Suggested Suggestions to Improve A/B Testing

Discussion in 'Suggestions' started by Tangerine Travels, Sep 2, 2018.

  1. Tangerine Travels

    Tangerine Travels Newbie Member

    Joined:
    Apr 20, 2018
    Messages:
    5
    Likes Received:
    3
    I have a few suggestions to improve A/B testing:
    - Allow us to easily add emojis in the "Title" section.
    - Allow us to extend the tests at any time (for the times when we don't have enough data to pick a winner)
    - Allow us to end the test early with the option to keep the variation (sometimes, after just a few days, the variation is already the clear winner)
    - After the test completes, give us the option to keep the variation OR revert back to the original, regardless of what the data says. Too often is the case where the original got more views, but all the other stats say "go with the variation." In this case, it'd be nice to be able to click a button that says "Change to the variation."
     


    Paul and Andrew like this.
  2. Andrew

    Andrew Superman Administrator

    Joined:
    May 6, 2016
    Messages:
    3,295
    Likes Received:
    2,217
    Pretty solid suggestions! Will send this to the team and tag @Phil
     


    Paul likes this.
  3. Tangerine Travels

    Tangerine Travels Newbie Member

    Joined:
    Apr 20, 2018
    Messages:
    5
    Likes Received:
    3
    Thanks @Andrew! I'd also like to see a watch time/impression statistic because this is something we think is one of the most important ratios for YouTubes algorithm. For example, a video gets 1,000 impressions at a 5% click-through rate which equates to 50 views. The average time viewed is 10:00 which equals 500 minutes of watch time for 1,000 impressions, or 30 seconds of watch time per impression.

    It would also be nice to see a character count and tag rankings in the tag section.
     


Loading...