MCPWhat an import does
- Scales. NPS is 0-10. CSAT and CES are 1-5 or 1-7, which you choose. A score outside the scale is skipped with a reason and never rescaled.
- Matching. GainTrace matches each respondent’s email to an existing Person, and failing that to the one Company whose email domain it is. If the Person belongs to several Companies, the answer is stored against their Primary Company. An email at a free domain such as gmail.com is never matched to a Company by its domain, so it matches only through a Person. The answer is stored against that Company, and against the Person when there is one. A respondent it can’t match to a Company is skipped with the reason “No Company matched the respondent email. Sync or import the Company first.” An import never creates People or Companies, so sync your CRM or import your People and Companies first.
- No duplicates. Each response is stored once for its tool and response ID. Pulling the same response again updates it if it changed and never duplicates it.
- Quiet history. Answers from before you connected, and every CSV import, arrive quietly. They send no webhooks, start no Flows and post no alerts. Answers that arrive after you connect send a
survey_response.createdorsurvey_response.updatedwebhook with the channelimport. - What they count toward. Imported answers appear in Reports, tagged with their source, and count toward Company scores, coverage, the frequency rule and the CRM fields GainTrace writes. They don’t appear in the Response inbox, which is for your own surveys.
Connect a tool
Choose Connect a tool. Owners and admins can connect and disconnect a tool. Name the connection, choose the tool and the measure, and fill in its details:
Optionally set Import history from to the earliest date to bring in. GainTrace tests the credential and the mapping when you connect. A problem it finds, such as a form whose score field has the wrong number of steps, is reported right there.
Rules for each tool
- Typeform. The score field’s number of steps has to match the scale you chose.
- SurveyMonkey. The score question needs 11 answer choices for NPS, or as many as your CSAT or CES scale. SurveyMonkey weblinks that don’t collect an email carry nothing GainTrace can match, so they import nothing. A SurveyMonkey Basic plan limits what its API returns to 25 responses.
- Qualtrics. Qualtrics prepares an export before GainTrace can read it, so the first pull can take a couple of minutes. The history date isn’t applied to the export. GainTrace downloads it and drops answers before that date afterward.
- Retently. If GainTrace can’t read the feedback list or the ID, date and score fields, connecting fails with a message that says which one.
Schedule and limits
GainTrace pulls new answers every 15 minutes for Typeform and Retently, and every hour for SurveyMonkey and Qualtrics once it has caught up. Typeform holds back responses under 30 minutes old, which arrive on the next pull. GainTrace stays inside each tool’s published rate limits: 2 requests a second for Typeform, 120 a minute and 500 a day for each SurveyMonkey app (it stops with 50 left and resumes after the reset), and 150 a minute for Retently. For each connection you can choose Pull now, Pause or Resume, and Disconnect. The row shows its status, the last pull, how many answers it imported and skipped, and an error if there is one.
Disconnecting deletes the saved credential and keeps the responses you already imported. GainTrace stores credentials encrypted and shows only the last characters.
Import a CSV file
Choose Import a CSV file, pick a file format (Delighted export, Retently export or Any CSV) and choose the file. GainTrace suggests a mapping from the column names. Always check it, because you can change any column.
Choose the Measure from a Survey type column in the file, or pick NPS, CSAT or CES for the whole file, and set the CSAT and CES scale. The sheet previews the first rows and tells you how many can’t be imported and why. Rows import in batches, and if one stops partway you can import the file again to finish without duplicating anything.
Re-importing the same file never duplicates. With a Response ID, that ID identifies the row. Without one, GainTrace identifies a row by the email, score, date and measure. In Any CSV, a response ID is combined with the respondent’s email, so two exports that both number their rows from 1 can’t overwrite each other, and GainTrace never picks an ID column on its own.
When it finishes, the sheet reports how many rows were imported, updated, unchanged and skipped, with the main reasons. Skip reasons never repeat a respondent’s email.
Where imports show
- Surveys > Reports and its By source view, tagged with the tool or
csv. - The person’s Feedback tab and Company scores.
- Surveys > Who you hear from.
- The Surveys API:
GET /api/v1/nps-responseslists imported and pulled NPS, and you can filter it byprovider. The API also lists your connections and can pull a connection now, and MCP has matching tools. Imported CSAT and CES answers appear in reporting and Company scores, not inGET /api/v1/survey-responses.