The number of votes, indicating interest, is one of the important inputs when evaluating idea priorities.
Currently, when an idea is merged, their votes and subscribers are merged to the primary - assuming duplicates avoided if there is a limit of 1 per person. The vote details (who) are also important to assess the coverage of those users to gain additional feedback, the number of companies represented, types of segments, etc.
However, if any child ideas had proxy votes, the proxy vote details don't carry over - they are neither added to the primary vote nor do the vote details indicate that any of the inputs are proxy (thus, appearing as if they same user voted more than once, even though they should be limited to only 1 - or whatever your setting is).
This requires that you navigate to each of the merged ideas to gather the information. It would also not show accurately on votes/proxy vote details by primary idea in any report.
Since this seemed to initially be a defect, I submitted through support - but, they indicated it would be an enhancement to have the vote consistency
We are rolling out a new idea portal and process to help in consolidating the ideas, making it easier for individuals to know where to go. However, because of our extensive user base, especially within larger customers, we are often inputing proxy votes after specific meetings based on the requests vs requiring those customers do the same. It allows those inputting to also provide more clarity and context on the need.
Because the proxy votes are not added to the merged vote, these are the other issues encountered:
Idea portal inaccurately reflects vote count on the Popular tab
Idea reporting based on votes and/or organization names are inaccurate and results in inaccurate trending or analytics based on reports or understanding of specific organization interest. The value of merge is to manage duplicates . However, you have to include all of the merged ideas to get insight, but when you have to include the merged ideas and their data, the report results become untenable to evaluate and analyze.