How do I submit a campaign for Persado to optimize?
You can request Predictive Content or Experiments any time by logging into Persado Portal at portal.persado.com and clicking the “New Experiment” button. A form will pop up where you’ll fill out some basic information. For a comprehensive walkthrough, please refer to this article: https://library.persado.com/en/articles/6989193-how-to-request-a-campaign.
What is the end-to-end process for running a Predictive Content or Full Experiment campaign together?
Running a Predictive Content or Full Experiment campaign occurs in a few simple phases.
Content Requested (Customer): Using Persado Portal, you’ll submit a request for your desired campaign. We’ll have regular status calls where you can ask us any questions if you get stuck! You’ll want to include the following information in your request:
The control creative
The desired channel for deployment
Your desired deployment date
Your primary KPI
Additional details on strategy/background.
Variant Generation (Persado): Persado uses AI to generate Variants (typically 1-4 for Predictive Content and 16 for Full Experiments) using best performing language, emotions, and concepts according to your brand, industry, channel, and campaign. We generate Variants in about 2 hours for text channels, and 6 hours for rich content, like a web page or email body, though times may vary depending on campaign type.
Variant Review and Approval (Customer): Using Portal, you’ll provide any updates required to the Variants. Persado will then incorporate these revisions in 3 hours for text channels (e.g., SMS, subject lines, push notifications) and 6 hours for rich content (e.g., full email, social, display). You’ll approve the final versions prior to deployment.
Deployment: Deploying Predictive Content and Experiments doesn’t require any integrations; we’ll typically follow your business as usual process for deployment like you would for any A/B test. If you do want to use one of our integrations, our Solutions Consulting Team can assist you with the ‘Setup’ tab in Portal.
Experiments only: Because Persado Experiments implement experimental design, a few more steps occur following the first round of deployment (called the Exploration Phase). Once the Experiment Report for the 16 deployed Variants is ready, Persado analyzes the results to determine which elements were the best performing. With these learnings, a Best Predicted Message is created to deploy in a second round (called the Broadcast Phase) alongside the control and 1-3 other Variants designed to maximize the lifts seen in the Exploration Phase.
Final Results: Following deployment, we’ll follow your standard observation windows for conversions. You and your CS representative may agree, for instance, to huddle up after the Variants have been in market for 1 week. Once the results are available, you’ll be notified that your Experiment Report is ready in Portal, where we can see how each Variant performed against your KPIs!
For more in-depth information, check out our Introduction to Full Experiments and Introduction to Predictive Content articles.
What campaigns can I test? How much volume do I need?
Persado currently supports most digital channels (email, web, push, SMS, social, adwords, and display).
When you’re running a Full Experiment, we recommend the following volumes - but they’re not required. These volumes help us reach statistical significance within 30 days, which gives us greater confidence that our results were caused by our updates to the language, and not by random chance:
For emails: 3,500+ unique KPIs in either open, click, or conversion for each send (or within 30 days for evergreen content/lifecycle content/triggers)
For Web/Mobile: 7,000+ unique KPIs in either clicks, applications, or units within 30 days.
Please note that when using Predictive Content, reaching statistical significance is only a requirement if the Predictive Content is placed into market against a control.
Do I always need to deploy a control in market?
We always recommend deploying a control in market if one of your goals is measuring performance (e.g., uplift, incremental impact versus a baseline). Deploying a control early on in our relationship is especially helpful as we can better measure the impact of your updated Variants. However, please note that including controls alongside Predictive Variants is optional.
Persado's Data Science team can also assist you with estimating uplift/incremental value in the absence of a control, which may be preferred in the following scenarios:
A winning Variant has already been declared and you prefer to not "waste" audience volume/media spend/resources on the underperforming control
You'd like to roll out a winning Variant on a new placement. You may think of this as using a Persado Variant “business as usual” (BAU).
You're using a platform that does not support comparison against a control. This can be for a variety of reasons, such as inability to randomize, or inability to report results per Variant.
There is no control Variant (for example, when you are running a campaign for the first time).
What kinds of reporting are available once the campaign goes live?
Once the status of the campaign in Portal changes to “Reports Available,” you can click into that campaign to see insights such as best-performing language, element contribution, and individual Variant performance, divided into various KPIs. The exact content depends on the type of campaign you ran, such as an Experiment, Predictive Content, etc.
For more in-depth information on reporting and insights, including aggregate data and industry insights, check out our article on How to View Results and Insights.
What happens after the campaign?
Once a campaign has gone live, you can consider the following next steps:
For evergreen campaigns like triggers, keep the winning Variant(s) in market to continue to drive performance (either with or without a control). You can submit a new request once you're ready to refresh that campaign (e.g., after 3 to 6 months).
For ad hoc/one-off campaigns, continue to request Predictive Content in Portal to access additional high-performing content.
Share the learnings and insights with your creative/copywriter partners so they can use it as inspiration. Your copywriting team might consider applying these learnings to areas where you aren’t testing with Persado, like new channels or brands. For more information on reporting and insights gained from campaigns, check out our article on How to View Results and Insights.
How do I manually share campaign results with Persado?
The easiest way to share results is to attach your files in the comments section for that particular campaign. (If you have an integration, your results will feed in directly to Portal.) Simply click the button that states ‘Click or drag files to this area to upload,’ and then click ‘Comment’ when you are done.
How do Variants approvals work with multiple teams approving?
When you have a need for multiple teams to approve, please use the new “Teams” feature in Portal. This will allow you to see which teams have approved particular Variants and when.
When using this feature, if one team requests revisions to Variants or makes a manual edit, approvals for all teams involved will reset to ensure they have an opportunity to approve the latest and greatest.