A company streams click data into Kinesis and delivers it to an S3 data lake via Kinesis Data Firehose. As volume grows, S3 ingestion rate remains steady but Kinesis Data Streams and Firehose backlog grows. Which action is MOST likely to increase the ingestion throughput into S3?

Increasing the number of shards in the Kinesis data stream directly increases its capacity to ingest and process data. Each shard provides a fixed unit of capacity for data input and output. By adding more shards, the Kinesis stream can handle a higher volume of incoming click data, preventing backlogs and allowing Kinesis Data Firehose to process and deliver data to S3 more efficiently. Decreasing the retention period would not increase throughput; it would only reduce the amount of historical data available. Increasing S3 prefixes offers no direct throughput benefit for Firehose ingestion. Adding more consumers using KCL would be relevant if the bottleneck were downstream processing of the Kinesis stream, not the ingestion into S3 via Firehose.