Once you’ve submitted a project, you’ll receive a detailed Compliance analysis flagging risk level, risk observations, alignment to brand voice, and predicted performance in market. This article will outline how to interpret this analysis as well as how to act on the feedback given.

It’s helpful to review the full analysis up front so you can make a plan for revisions. Starting on the right side of the Analyzer, you’ll see all of the observations found in your creative that might pose compliance risks. The red observations are considered higher risks, and orange observations are lower risks. If any content is not marked up here, it has passed all of our compliance checks!

You’ll see a summary of different risks at the top of the page, with numbers indicating how many risks were flagged in each category (e.g., 2 priority observations for omitting conditions). If you’d like to review all of the resources that our AI checked for, click the ‘All Resources’ tab for a refresher of what we’re analyzing against, and where potential risks were flagged.

Check out our cheat sheet below on how to interpret compliance summaries.

All of the risks listed under priority and secondary observations are calculated into the overall risk level score shown on the left. The lower the score, the better! However, you won’t be able to resolve certain risks in our interface, so it’s okay to submit copy with a certain amount of risk to compliance. Let’s go through some of the findings so you’ll see why.

Next, you’ll go through each observation and take action. Click on any observation to expand it and reveal its details.

Let’s go through a few observations so you can see different options for taking action on compliance risks. Depending on the observation, you’ll be able to:

We recommend pulling up the cheat sheet below for your first few reviews!

We recommend reviewing priority observations first, as these are considered by our model to be high-risk findings. Some compliance themes are very likely to be marked as priority (e.g., credit deception), but ultimately the severity depends on how a regulation is violated. Each observation will come with this context.

Let’s take a look at this priority observation around the omission of conditions. Here, our model is flagging that FCRA prescreened offer disclosures are missing and includes suggested disclosure language that should be added to the full email.

You can manually edit disclosures or any area within the creative within the Analyze tool. Or, once you make the change offline, you can mark risks like these as “Resolved” and add that explanation, and move to the next observation. You’ll be prompted to provide an explanation when:

Observations may also come with recommended revisions that you can incorporate with one click to resolve the risk. Hovering over the quoted text to view the language our model is flagging will help you select the appropriate revision.

In this example, you can click the blue “Use” button to address the “Guarantee” concern that the copy is making a definitive promise about the impact of the card on a user's financial future. When you’re happy with the update, click “Resolve.”

Secondary observations are still important to review and assess with your team, but depending on your experiences with your own Compliance team, you may know that certain language is okay to use. In these cases, you can resolve the observation, which will help train our model to understand the unique needs of your brand.

For example, you may know that using language like “Exclusive Elite Awards” has passed Compliance review in the past, so you could mark this as “Resolved.” Note that for audit purposes, you’ll need to provide a rationale as to why the status of the risk has changed.

Performance Observations work a little bit differently than Compliance Observations in that they do not use severity levels (like ‘priority’ or ‘secondary’). Instead, every alternative displays:

To swap in an alternative, click on the check mark as usual. Scores are recalculated automatically as alternatives are applied; reanalyzing is not required for score recalibration. Then, if you’re all set, resolve as usual.

Performance-Only Observations are shown with a blue outline and turn green when resolved. They will be clearly marked to indicate that while there are no compliance risks detected, our AI does see an opportunity for general improvements to the copy—like in the “Apply Now” CTA below.

Incorporating updates for Performance Observations is even simpler: just click the blue check to update the copy, and the language and overall performance score will automatically update.

You may also opt to make manual edits in Portal, whether it’s in response to an observation or just updating the general creative. Perhaps you’d like to combine our recommended revisions, update disclosures to address risks, or make a change you know has worked well in the past. For any reason, you can click the edit button (i.e., the pencil icon) or double click on the text box of the creative to make tweaks in Portal. Hit ‘Enter’ when you’re done, or click your mouse outside of the text box (but within the creative), and the change will be saved. If you’d like to add a new line of text, use the Shift+Enter shortcut.

Or, you can leave thoughts for your teammates using the comment button. When you’re happy with the updates, mark “Resolved” as usual.

As you make changes to reduce risk, click “Reanalyze” in the upper righthand corner any time to receive an updated risk score and compliance summary. Brand Alignment and Performance Prediction scores will be updated in real-time as you make edits; you can learn more about how to factor these scores into your analysis here.

When you’re done, you should see all observations resolved and a green check indicating that our model doesn’t detect any compliance risks. While it’s not mandatory to resolve all compliance risks noted, we recommend that you resolve as many as possible before submitting for final review. Your threshold for risk will be up to you and your team!