Rating, Evaluation and Testing: QA flow for negative feedback/complaints on translation tasks.
Software rating function
On the Software platform customers can leave a translation rating from 1-5 on each task. Ratings between 1 and 3 are considered negative, 4 and 5 are positive. If customers leave a rating of 4 or 5 stars the linguists receive an email and the score will be tracked in HubSpot without creating a ticket. For ratings from 1-3 the system will trigger a complaint ticket in the complaint pipeline in HubSpot.
Example ticket: https://app.hubspot.com/contacts/8932080/ticket/1081474931/
This customer feedback will be sent to both Vendor Management and the respective translator, so they will already know the comment. When you receive a negative rating please go to the ticket and see the customer’s feedback.
Example:
On the top right of the ticket you see who uploaded the project and which translator was working on the task:
Evaluation
1. Start out by getting in touch with the respective CSM and make them aware of the rating (they will also see it in the #negative_rating Slack channel) and discuss if you should take anything into account when speaking to the customer.
2. Reach out to the customer (use the ticket title as the email subject) and let them know that we received their rating. Depending on the feedback, ask them if they would like the linguists to amend the translation, if they could share some more detailed feedback or just let them know that we are on it.
Example:
3. If needed ask the Support team to reopen the task.
4. Reach out to the translator and let them know how the customer would like to proceed with this issue (e.g., reworking the task, taking feedback into account for future tasks, etc.).
Quality check
If the customer mentions severe issues with the quality of the translation, we should perform a QA on appr. 500 words of the task. For this occasion, please assign a formal QA on the Service Platform as you would do for a simple LQC tasks.
Find the files for the evaluator here.
Find more information about formal QAs here. Please make sure to assign suppliers with QA experience: Evaluator List
Tracking
Once you have investigated the problems, you need to track the nature and category of the complaint - these are the relevant Complaint Management properties:
Most of the choices are self-explanatory. If you don’t perform a QA, choose “Negative feedback”. If you need to warn the translator, choose “Warning given”.
If you do a QA, remember to log the result under the associated deal, so it’s connected to the project ID and the translator:
Go to the deal and find “Supplier QA score” under the properties on the left side:
Please leave notes in the tickets if QAs are done and if any action was taken that cannot be read out from the emails associated with the ticket. When we go back to the ticket in half a year, we still need to be able to understand the context.
In general: Make sure to communicate with the respective CSM/the Support team. Quality issues need to be assessed on a case to case basis as a lot of factors are playing in. If in doubt, ask for help from your colleagues and your manager :-)
Possible consequences for translators
- formal warning
- remove translator from customer team
- offboard translator from platform
- payment deduction
The Complaint Management is part of the Supplier Monitoring Protocol. Please check it out for details on how to issue warnings and how to document general quality issues.
In general, it is important to distribute learnings from evaluations to other linguists in the team. If general issues arise, make sure to share information to all involved linguists as a general heads up.
Overview of Complaint/Quality process:
In the top row you find the steps from the HubSpot pipeline.
Below you can follow the steps for the standard quality processes.