Begin by analyzing the data structure within Tempo. Identify the types of data, format, and fields you need to extract. This understanding is critical to ensure compatibility with MongoDB's document-oriented storage model. Document the schema and any transformations required to map data fields appropriately.

Develop a script or use Tempo's native export functions to extract data. This could be done using REST API calls if Tempo provides such an interface. Ensure that the data is exported in a structured format like JSON or CSV, which can be easily transformed and imported into MongoDB.

Once you have the data extracted, you may need to transform it to ensure compatibility with MongoDB's BSON document structure. If your data is in CSV, convert it to JSON. Use a programming language like Python, JavaScript, or Java to write scripts for any necessary data transformations.

Prepare your MongoDB environment by creating the necessary databases and collections where your data will reside. If you haven't installed MongoDB yet, download it from the official MongoDB website and follow installation instructions suitable for your operating system. Use `mongo` shell or MongoDB Compass for database and collection creation.

Write a custom script to import the transformed data into MongoDB. You can use Python with the `pymongo` library or Node.js with the `mongodb` package. The script should read the transformed JSON files and insert them into the appropriate MongoDB collections.

Run your import script to transfer the data from the extracted files into MongoDB. Monitor the process to ensure all data is correctly inserted. Handle any errors by logging them and rerunning the import for failed data entries if necessary.

After the import, verify the data integrity in MongoDB. Use queries to check if all records are present and correctly structured. Compare a sample of records between Tempo and MongoDB to ensure data fidelity. Make any necessary adjustments and rerun the import if discrepancies are found.