# Topic Design Best Practices

**URL:** <https://community.hivemq.com/t/topic-design-best-practices/615>\
**Category:** MQTT\
**Created:** [July 1, 2021, 1:39am UTC](https://community.hivemq.com/t/topic-design-best-practices/615 "2021-07-01T01:39:46Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![ThePopeOfAntelope](https://avatars.discourse-cdn.com/v4/letter/t/73ab20/32.png) [@ThePopeOfAntelope](https://community.hivemq.com/u/ThePopeOfAntelope)\
**Post date:** [July 1, 2021, 1:39am UTC](https://community.hivemq.com/t/topic-design-best-practices/615/1 "2021-07-01T01:39:46Z")

</div>

Our clients want telemetry data from our buildings via MQTT. We have 10+ buildings and 100,000~ data points. We will double in size in 5 years. Our devices can publish MQTT. I would like to aggregate data with HiveMQ and filter data for clients based on devices that monitor their inventory using a plugin. Each device has 5-10 data points. I’ve seen examples where 1.) a topic is used for each data point and 2.) all data points are part of a single topic in a JSON format. In either option I can filter on content/Topic to reroute data to the correct client. Which is considered the best practice a topic per data point or a topic per device based on the current/future number of devices? Are there other things to consider? Has someone published a paper on this?