Just finished migrating our lead-scoring models from Marketo to HubSpot. The built-in tools were too slow for our dataset (~50k records with complex scoring attributes). So I rolled up my sleeves and built a custom Python script.
Used HubSpot's API for the write operations and a direct MySQL dump from Marketo for the source data. The trick was mapping the custom score fields and batch processing to avoid rate limits. Had to handle some data cleanup on the fly, like normalizing lead statuses. The whole process took about 2 hours for extraction, transformation, and loading. The actual cutover was just flipping the integration off and running the script. Anyone else done a similar migration? Curious about your approach for handling custom field logic.