The state of Kafka replication:

60% of enterprises are neutral to dis-satisfied with the Kafka replication technology they rely on. This is despite the fact that disaster recovery is cited as the primary reason to replicate Kafka data, meaning that what engineers depend on in a crisis is also where they report the least confidence. 

This report asks 150 practitioners running Kafka at scale how they can trust what arrives on the other side, catch problems before customers do, and trust the tools they are locked into commitments with:

Is more headcount fixing the issue, or absorbing it? 43% of teams put three or more teams on replication, and 84% of them still need a week or more to stand up a single job. 

Is the tool your team likes best also the one boxing you in? The highest satisfaction platforms carry the highest lock-in anxiety – a tension that sharpened the moment IBM closed its $11B acquisition of Confluent. 

If you’re replicating for disaster recovery, can you trust the failover? 71% of DR-driven teams cite broken offset translation or blind observability; the two things a failover depends on.

Does your risk function know what your engineering team already suspects? In financial services and healthcare, schema drift and masking gaps during replication are potential sources of compliance and audit risk.

Will your infrastructure hold up as AI workloads scale? 67% say AI/ML growth is already increasing replication demands, and training and inference pipelines have no room for the duplication and lag current tools allow.

If 70% of the market is ready to switch tools, what's holding you back? If you lead engineering, platform, or risk at an organization running Kafka at scale, this report shows you where your exposure sits and what constitutes a switch worth making.

Watch the webinar with Guillaume Ayme, CEO of Lenses.io, and Dinesh Chandrasekhar, CEO of Stratola Research.