Some aggregations on SPARQL endpoints can be quite slow, yet these can produce very interesting visualisations. It would be great to be able to cache the query results so that the data story will load quickly. This of course means the data is not live, so it should be an option and it should be possible to refresh the results if necessary.