You could have a look at the following search (unless you've already solved the problem yourself). It makes use of the field date_month which is calculated for all events (apart from any windows event logs), along with date_hour, date_minute, date_wday, etc etc.
your_search_here | chart sparkline count AS usage_count over userid by date_month
userid sparkline:september usage_count:september
adam graphics here 43
bill graphics here 92
cecilia graphics here 238
As you will see, it does not edit the actual column header, but rather uses the date_month as its header, of sorts. However, it will only look as intended if you limit the search so that only events from a single month are returned. The search will work for other time-ranges as well, but the output will not be what you desire.