Wednesday, February 12, 2020

A Rating System for Ancestry's Public Trees?

Judging from what I see on social media and other internet forums, the wider genealogical community relies heavily upon's public member trees for research clues and information. Despite the copious warnings about the general status and reliability of these trees that genealogy pundits issue to amateurs, I don't see this dependence changing any time soon. So I've been thinking: what if there were some kind of rating system in place for Ancestry's public trees? What could it look like? Why would it be desirable?

Beginning with the last question, it might help researchers of all skill levels to see at a glance which public trees are particularly well-sourced, organized, and otherwise maintained. Trees given special quality designations might quickly attract information/clue seekers to the most helpful places. They would also serve as models for users who are looking to improve at genealogy, and would show them what good work looks like specifically at this most popular of genealogy sites. Such a rating system might also motivate many public tree maintainers to be the best they can be by pursuing marks of distinction for their trees, which may in turn provide them with the widest possible exposure.

What would such a system look like? I believe that a few factors could (and probably should) be considered. I have some ideas below.

1. As with the current choice to make one's tree public or private, a rating system could be entirely optional. If users don't want to have scores/badges/whatever attached to their trees, they could opt out. Or, they could opt not to publicly show ratings that get assigned to their trees. Would this limit the amount of public trees that show ratings? Perhaps, but I do think that a fair number of people would opt in. It doesn't have to be millions, but thousands or even hundreds of takers may still make the venture worthwhile.

2. One of the things I like about the Rotten Tomatoes movie/tv show reviews website is that it provides both critical and audience scores for each item. An Ancestry public tree rating system could work similarly. Users could provide star ratings for public trees they use, and trees highly rated by the community could display a special badge or sticker. Genealogy professional/expert ratings for public trees could work the same way. (Such users would have to be identified and given special rating privileges.) There wouldn't have to be "negative" tree ratings displayed. If trees receive lower scores, they simply wouldn't have badges or stickers to show. This ensures that nobody is negatively singled out or shamed.

3. The rating system itself could use a 5-star system like Rotten Tomatoes and GenSoft Reviews do. Tree users could answer optional questions about a tree's sourcing, detail, and accuracy as they perceive them. A composite rating could then be assigned based upon the marks from these answers. A community score and an expert score could be presented side by side for each tree. This gives a tree's potential users a very good idea of its quality or usefulness.

4. Since family trees are works in progress, there could be expiration dates on poor reviews to allow tree maintainers to make subsequent improvements. Dates could be assigned to positive badges/ratings as well to allow for the possibility of declining standards in trees. The time element here is a bit tricky, I admit, but I'm confident that this problem is surmountable.

5. You could also simply have a feedback section for each public tree, GenSoft-style. Tree maintainers could select the best reviews and put them on a sort of front page for their tree, while the other reviews wouldn't have to be publicly displayed. This would be a very simple system, but it would at least increase community engagement and showcase good work. And again, this feature could be toggled on and off.

6. Maybe just a pool of experts, professionals, and/or Ancestry admins could give out badges to especially good public trees they see. Maybe feature testing would reveal that inviting the public into an expansive review system is impractical or would lead to negative experiences.

I don't expect this idea to be adopted any time soon, if at all. But even if Ancestry doesn't like it, they may want to consider other options that increase tree functionality on their site, especially at a time when this company (among others) finds itself facing financial headwinds and pressure to keep subscribers coming back. If Findagrave has shown us anything, it's that users like the game aspect of stats and achievements. Perhaps a public tree rating system is a way to harness those appetites in service of constructive ends.


  1. Something that might work, at least for me, is a "Tree Score" based on:

    1. What percentage of tree profiles have at least one source citation (not to another tree)?

    2. What percentage of tree events (all profiles) have a source citation (not to another tree)?

    3. What percentage of tree profiles have at least a birth and death date?

    Say that a tree has 50% of #1, 40% of #2, and 60% of #3. The average is 50%. If Ancestry ev er did something like this (and they do have the computer power to do it - it just takes a will to do it), they could list the trees with a specific person by the ranking.

    I'm sure we can think of other measurements if we work at it.

    Cheers -- Randy

    1. Good ideas. I'd love to see anything along these lines.