For most brands, weekly runs are correct. Weekly runs show a real change in some days. They are also not so frequent that the noise is larger than the signal.

​
Why not daily runs

Answer engines give different results from run to run. Most changes from day to day come from the medium, not from your market. When you monitor daily, you learn to react to changes that have no meaning. Daily runs are useful in two conditions:
  • In the week after a large launch.
  • When you test whether a change had an effect.

​
Why not monthly runs

In one month, a competitor can publish content, get citations, and take a position before you see it. When you see the change, it is difficult to find its cause.

​
The real limit

Your plan sets the number of conversations that you can run. Thus, the question is usually not “How often?” but “How do I use the budget?”:
  • More topics, less frequently. This gives breadth. It is better when you make a map of a category.
  • Fewer topics, more frequently. This gives a reliable trend on the important questions. It is better when you work on a specific problem.
Start with breadth while you learn the structure of your category. Then make the set smaller.

​
A practical pattern

Run conversations weekly across your monitored topics. Add a run after each event that you expect to change the numbers. Compare runs of the same type. A run after a publication has meaning only when you compare it with a run before the publication.