I couldn’t stand it any longer! I was at RootsTech 2019 a couple of weeks ago and decided to approach representatives from Ancestry.com. I actually spoke to four different people, hoping that perhaps I would receive feedback from at least one of them, but that hasn’t happened yet.
I am a pretty easy going person and have quite a bit of patience. I’ve been waiting for Ancestry’s search engine to improve instead of getting worse, but that hasn’t happened yet either.
I was prepared for my chats with the reps because I knew they were going to ask for examples of my issues. Here is one that I shared with them and it’s a biggie.
One resource that I needed for a blog post was the 1860 U.S. census slave schedule for Bourbon County, Kentucky. The person in question was Jacob Spears, who was an owner of enslaved people.
I entered just the surname SPEARS in the search box and specified Bourbon County, Kentucky and got this result:
No Results!
Before I realized how big a problem this was going to be, I tried spelling variations like Spear, Speer and Speers, also with negative results.
I knew from probate records that the family was in the county and thought perhaps they had been missed in the slave schedules. Therefore, I went to the 1860 U.S. federal census and looked for the family. Here is the first page of results – note that I specified Bourbon County, Kentucky and not one of these families is in Kentucky!
Okay, so maybe the family got missed by a census taker who was completing both schedules at once and actually did pass by the plantation without enumerating the household.
I tried a second surname, out of curiosity – Talbott. Back to the 1860 slave schedules for a new search. My results:
I looked for Talbotts in the regular 1860 schedule, too:
Nothing, nada, zilch! Not a Talbott to be found anywhere in Kentucky in the first hits!
Then I tried a third surname – Grimes:
No results!!! Each of these families had a family member who died in this time period leaving wills mentioning enslaved people.
A search from the regular 1860 schedule brought these results:
At least there is one Grimes hit for Bourbon County, but I happen to know that there are at least 3 different Grimes entries in the Eastern District alone.
Also, when I specified Bourbon County, why are there families from Georgia, Indiana, South Carolina, Tennessee and Maryland all coming up????
Were pages missing from the Bourbon County, Kentucky census? To figure out if that was the case (not wanting to read Bourbon County page by page if I didn’t have to), I went to FamilySearch and entered a search for SPEARS with a residence in Bourbon County, Kentucky from 1860 to 1860.
What appeared?
Jacob Spears is the first family enumerated on page 681! The family is in Bourbon County and WAS NOT missed by the census taker. I went back to Ancestry.com and this time went to the census browsing tab.
The manual page-by-page search on Ancestry.com found Jacob Spears:
However, the family was definitely missed by the Ancestry.com search engine.
Back to the Spears in the 1860 U.S. slave schedule. Knowing that the family lived in the Eastern Division, I went to the census browsing tool in Ancestry.com.
Jacob Spears found! Again, the search engine did not find him.
I went back to the Grimes and Talbott names and found both names in the 1860 federal census and in the 1860 slave schedule. In fact, the Talbotts were a good-sized family and I found the name on multiple pages in both censuses.
My conclusion is that this census is essentially UN-INDEXED!
Unfortunately, this problem isn’t confined to just Bourbon County, Kentucky. More, tomorrow.
You piqued my curiosity. I tried a search of Jacob Spears in the 1860 US Census at Ancestry with residence location of Kentucky and he (and other Spears) appeared in the results. The problem is when you want to narrow down to a specific county, which I don’t do anymore because I don’t get the results I expect.
Looking forward to your other example(s).
Cathy also pointed out that entering the place in the keyword search box works. It’s nice to have an alternate plan, but I am still hoping Ancestry fixes the place field so that it works correctly all the time.
When I was with the Ancestry rep and he searched for Spears in the 1860 slave schedules, I heard a quiet comment about there might be a major issue with the search engine. I find Ancestry to be almost worthless for the census records. I have better luck using Search All Collections and hope that the results find people in the census I want.
Ouch. This is bad news. Eagerly awaiting your next post.
I have a similar issue. My solution may also help you. First the problem: the search engines will not turn up counties which later became part of West Virginia in the 1850 and 1860 census when I search by county location. As their search engines only recognize these counties as being in West Virginia which was formed in 1863.
I find by using Virginia for the location where they lived and the name of the county (Fayette, Mason, Kanawha) in the Keyword box will pull up a list of persons from the county I am trying to search in. I tried this for Spears in Kentucky with keyword Bourbon. The first 22 hits were in Bourbon, KY for Spears with Jacob on #3.
Cathy, Your tip worked like a charm! I’ve been playing with it this morning. At least there is a workaround to this issue, but Ancestry still needs to fix the problem. Why have a Census search field for PLACE when it doesn’t work correctly???
I found Jacob Spears on Ancestry within one minute. It’s the same page you found on familysearch.org.
I spent quite a while talking to a young man from the Ancestry tech team while at RootsTech. I gave him these same examples and, believe me, he didn’t get any results when searching for the same items either. I know someone somehow fixed the Oldham County filmstrip of the 1840 census and I am actually hoping that this much bigger indexing issue gets fixed, too.
I’ve been an Ancestry.com user for over 10 years and their search engine has def degraded. I try simple searches, within specific collections, and many times I get no results, knowing that there are def records to fit the criteria in the database. It’s been very frustrating. I’ve reported a few, with details and get no feedback. Why are they ignoring this problem?
I wish I knew why they were ignoring these issues, but I think it is related to the fact that they are much more interested in selling DNA kits.
I doubt Ancestry will do anything about the search. They’re too involved with making money from DNA testing. I have called several times, even have a ticket number, but I’ve never received a reply, although the person I spoke to agreed there was a problem. I called them because the initial search results page will indicate 303 results for a specific database. Clicking on that database link will give me 175 results. So what happened to all those other supposed results?? Is the 303 number incorrect? I always search very broadly initially, so I should receive results based on ‘Soundex, Sounds Like, Similar.’ These parameters don’t seem to work much of the time, especially Soundex, which I have tested on people I had found years ago. I knew they were in New Jersey, but the current search, using just Soundex, did not bring them up.
Hi Dee Dee, It has been very frustrating watching Ancestry go downhill over the last few years, but I was told at RootsTech that they do read all the comments on their Facebook page. Maybe making the issues public like that will bring more results? We should all try it.
Linda,
I’m really glad that you posted screen shots of your search queries. URLS of your search results would have been better because then we could see exactly what your query was but, based on what I can see here, I think this is what is happening.
It appears that you have typed Bourbon, Kentucky into the LIVED IN field instead of selecting Bourbon County, Kentucky, USA from the type-ahead list. When you enter your own location instead of selecting from the list, the search engine treats it as a keyword and because you don’t have it marked exact it is somewhat downweighted in the search results.
When I select Bourbon County from the type-ahead (and then use the filters to mark it exact) and type in Spears in the surname field and mark it exact, the search result for Jacob comes right up.
https://www.ancestry.com/search/collections/1860usfedcenancestry/?name=_Spears&count=50&name_x=1_1&residence=_bourbon-kentucky-usa_349&residence_x=0-0-0_1-0
As for the the 1860 Slave Schedule, the problem you are experiencing there is due to the fact that the record is for the enslaved individual not for the slave owner. Ancestry created a record for the enslaved person listing their gender and age but, of course, there is no name. Technically, the name field should probably be removed from the search options. In order to bring up a list of enslaved individuals owned by a particular individual, you need to enter their name in the KEYWORD field, since they are not considered the primary person on the record. And, of course, the same location rules as above still apply.
So, a search in Bourbon County (using the type ahead and marking it exact) with a keyword of “Jacob Spears” (in quotes and also marked exact), returns exactly what you are looking for.
https://www.ancestry.com/search/collections/1860slaveschedules/?count=50&keyword=%22Jacob+Spears%22&keyword_x=1&residence=_bourbon-kentucky-usa_349&residence_x=0-0-0_1-0
I hope that was helpful.
I have a lot of issues with ancestry’s search engine.
Unfortunately, or fortunately(?), it is still one of the best places to track down historical data on people.
For me, the best tactic is to just keep tweaking the data I put into my searches. Sometimes less input into a search works better. e.g. I’ve even found that sometimes leaving a person’s middle initial off from a search is more likely to give results with their complete name (Which seems very strange!).
It seems to me that the voting record dates & locations are often listed wrong. for example: A San Franscisco 1922 voter registry is listed as one from San Diego in 1940.