General FAQ
Last updated
Last updated
If you want to understand the differences between various recommendation scenarios, we recommend checking out this article. The article provides a comprehensive overview of the seven recommendation scenarios offered by Rosetta.ai.
If you are using the right-edge tab for gift recommendations and want to change the icon, you can edit your Recommender. Click on the "Design" in the Advanced Options, and you can replace the icon for the gift box by entering the image URL in the designated area.
In Performance Indicators, "Total Order Value (GMV)" calculates the total value of orders from customers who interacted with and purchased recommended products.
The Engagement Value shows the total order value for customers who interacted with the recommendation but didn't make a purchase after viewing the recommended products.
Due to different sets of purchased products being considered, the numbers may appear differently.
If you find that the number of purchases from recommendations is very low, you can check based on the following three steps:
You can review the performance data of your individual recommendation: optimize the recommendation with lower CTR. How to check CTR performance?What to do if CTR performance is poor?
Check the changes of traffic in recent months: if the traffic is not high for previous months, the system may not accumulate enough data to predict your customer preferences accurately, which could result in recommendations that are less accurate or below expectations. How to increase the traffic?
Check your overall conversion rates: If the entire site's conversions are low, it may be necessary to optimize the consumer shopping experience.
The homepage doesn't need to be too long. It's best to place promotional messages in a prominent location, such as the large banner.
Ensure that the website category helps consumers quickly find the categories or products they are looking for.
The product page descriptions should be clear, as these are crucial in helping consumers decide whether to place an order.
It is recommended to use SEO to increase organic traffic
Use advertising to boost paid traffic.
CTR above 5%: When the recommendation's CTR is above 5%, means the recommended products align with customer expectations and they had high intention to click. The recommender are likely well-suited to your target audience's preferences which can be used as a reference to optimize recommender on other pages.
CTR between 2.5% and 4.9%: This range is considered normal for recommenders performance that the most recommended products meeting your customer expectations. It is advisable to maintain the placement and recommendation logic for this recommender.
CTR below 2.5%: If the performance of this recommender falls below expectations, it's recommended to adjust the position or recommender type and observe for two weeks (as data takes time to accumulate) to see if the CTR performance improves.
If you find that the CTR performance from recommenders is below 2.5%, you can follow these two indicators to check:
Visibility: You can check the visibility of individual recommender placements to determine if adjustments in placement are needed. Recommendations with visibility below 50% trigger a warning, indicating that consumers have limited exposure to the recommendations.
Position placement: Check whether your recommendation are placed in a right or ideal position
If recommenders are positioned at the bottom of the page, consider moving them higher as placements higher on the page typically have better visibility.
If the recommendation placement is already at the top of the page but visibility is still poor, it is recommended to observe whether customers can complete their shopping and transactions at the top of the page (for example, for products like underwear or footwear).
If the recommenders are presented as a "gift box" and require consumer clicks to be viewed, it's advisable to consider switching to embedded recommenders or pop-ups.
If you have good visibility for your recommenders but the CTR performance is still below expectations, it might be due to the recommended products not meeting consumer expectations. In such cases, you can optimize and test the recommendation type. (It's advisable to test each recommendation logic for at least two weeks for more accurate data.)
You’ll Love Me: AI-powered recommendations that utilize consumer clicks and popular in-site products to make recommendations.
Trending Now: Products that have been popular in the last two weeks (considering both clicks and purchase data), which may better align with the overall consumer buying preferences on the site.
Effective from December 1st, 2023, existing users in Taiwan area can change their default billing currency to NT Dollars by following these steps:
Make sure the currency shows as "NT$", then click "Select" to resubsribe to the current plan with new currency.
Completing the above will switch you to an NT Dollar subscription!
Please Note: Cancelling the subscription will prorate and refund any remainder based on your original subscription term, before charging the newly selected plan in NT Dollars.
When a customer enters a web page with Rosetta AI recommendations, the system automatically generates a new list of recommended products and recommendation positions. Each generation process is counted as one instance of recommendation impressions count.
You can start by identifying the pages where you want to place recommendations and adding up the monthly traffic on these pages. You can estimate the recommendation display count based on traffic of these pages.
Choose the best recommendation pages for more guidance.
Click Write a request email to our CS and the system will open your Email app with your email address and a template pre-filled. Please fill in and send the email to contact us, telling us the reason for cancellation and any other feedback. The pre-filled content is:
Recipient (Rosetta.ai CS Manager)
Subject line: Rosetta.ai Contract Cancellation Request
Email body - You need to fill in:
Store name: Type your store name
Cancellation reason: Type your reason for cancellation
Feedback to improve our service: Type your feedback on our service
Send the email.
Go to Account Settings > General Settings and to Traditional Chinese
Click the Cancel Subscription button on your plan details page (check for cancellation instructions).
