Skip to main content

Providing Feedback via Text Selection

Learn how to specify your content preferences to guide our AI.

A
Written by Amy Blakemore

Our Optimize Studio capability offers increased flexibility for providing Content Feedback!

With Content Feedback, you can now specify your content preferences simply by highlighting text and clicking 👎 (Less Like This) or 👍 (More Like This). This quick and lightweight feedback can guide our AI to provide content your customers will love.

Need more information? Let’s check out some best practices below.

Content Feedback Best Practices

With Content Feedback, you’ll be able to provide feedback to specific words and phrases that appear within a single content generation. This means that:

  • You can highlight the word or phrase you’d like to provide feedback to and suggest more (👍) or less of it (👎).
    ​

  • Once you log this feedback, it will be summarized for your reference on the sidebar of the particular campaign you’re generating content for. Keep in mind that the feedback you provide will be specific to this generation—that is, when you click ‘Generate,’ you’ll see more (or less) language based on your feedback. In the example below, you would see more mentions of “exclusive.”
    ​

Brand-wide Feedback

You may be wondering: how can I provide feedback not just to a specific campaign, but to my overall account?

In the Content Feedback sidebar, you’ll see the option to turn your feedback into Brand-wide Feedback. This means that what you’d like to see more or less of will apply not just to the specific campaign you are working on, but rather to all campaigns you generate content for in Portal.

While this feature is powerful, we recommend only adding to Brand-wide Feedback in the following scenarios.

Brand-wide: "More like this"

Applicable when you see words or phrases that are on brand, creative, or have performed well in the past, and you would like to see more of this phrase in every type of campaign going forward.

Example Scenario

You are a retail brand running a campaign about a spring sale on dresses, and you've generated the following outputs:

You have found that the phrase “must-have” always resonates with your audience, and you would like to always see options that incorporate “must-have” in every single campaign going forward.

You may add this to your brand-wide feedback, and you can expect to see “must-have” much more frequently in any campaign going forward.

You can see here that “must-have” is incorporated in an entirely different campaign for Black Friday.

Brand-wide: "Less like this"

Applicable when:

  • This feedback is relevant to every type of campaign my brand runs (e.g., promotional, trigger, seasonal).

  • The selected text is completely disallowed language - for example, because it will not pass legal requirements or the language makes claims that are misleading in any way.

  • The selected text does not align with your company's core values and messaging guidelines.

Example Scenario

You are a financial services brand running a credit card campaign, and you see the following outputs:

As a brand policy, you don’t allow the word “Congrats” because it is not the correct tone for any of your campaigns, and do not want to see it appear in any campaigns going forward. In this case, we suggest adding this word to your brand-wide “less like this” list.

You will no longer see “congrats” appear in any campaigns going forward, even if “congrats” is a high performing term or if “congrats” is applicable for a certain campaign you are running.

Did this answer your question?