Uncovering The History Of Obitmichigan Com A Rare Treasure Trove Of Vintage Obituaries
The State of Michigan maintains a massive archive of death notices going back to the late 1800s, and the primary public-facing portal for many researchers is obitmichigan.com. I have spent hundreds of hours digging through these records looking for specific families in Kent County, Muskegon, and the Upper Peninsula, and there are a lot of things that make this process significantly more difficult than you would expect from a simple search box. Let me walk through how it actually works and where the records break down. The site pulls data from multiple sources. Some entries come directly from funeral homes that opt into the system. Others are pulled from newspapers that have digitized their archives through subscription services like Newspapers.com or local historical society partnerships. There is also an older database maintained by the Michigan Department of Community Health that feeds into this, which contains certified death records from around 1921 onward. The exact provenance of any given entry is rarely labeled clearly on the page itself. You have to look at the formatting of the data fields to figure out which source you are dealing with. When you search for someone, you will get a list of results ordered by relevance. The relevance algorithm is fairly basic. It prioritizes exact name matches and then sorts by date proximity if you provide a year range. The problem is that many obituaries from before 1950 do not include complete dates. They will list just the month and sometimes skip the day entirely, or they will use an approximation like "circa 1893." This makes date-range filtering unreliable for anything older than about 1940. I learned this the hard way when I was searching for a great-great-grandmother born in 1878. I kept narrowing my date range and getting zero results, when the actual record had a typo in the year field that listed it as 1879. I had to abandon the search box entirely and browse by county instead, which is a feature the site buries pretty deeply in the navigation menu.
The county browsing option is your best friend if you already know where the person lived. Select a county, then select a year, and you will get a paginated list of every obituary from that county in that year. Each result gives you the full name, date of death, and age at death. Clicking on the entry takes you to a detail page that shows the full text of the obituary if it was published by a newspaper, or just the basic certified record if it came through the state system. The detail page does not indicate which source it is pulling from. That is something you have to infer based on the length and formatting of the text. Newspaper obituaries tend to be longer and include family member names, church affiliations, and service locations. State death records are shorter and more clinical. They list the cause of death, the attending physician, and the burial location. One thing that trips up a lot of people is the spelling variations in names. Before the 1960s, spelling was not standardized across different records. A single person might appear as "John Smith" in one obituary and "Jon Smyth" in another, even within the same family. The search function does not handle wildcards or phonetic matching. If you misspell the name by one character, you get nothing. I got around this by keeping a running spreadsheet of every spelling variation I encountered for a given family name, then searching each variation separately. It is tedious, but it beats staring at a blank results page. Another issue is the coverage gaps. Not every county in Michigan participates equally. Some rural counties have very sparse coverage, especially for the period between 1900 and 1930. The reason is mostly economic. Digitizing old newspaper archives is expensive, and smaller papers do not always have the resources to partner with commercial aggregators. If you are researching someone from a place like Ontonagon County or Houghton County in the early 1900s, you should expect significant gaps. I ran into this when trying to track down a relative who died in 1912 in Menominee County. The state death index had the record, but the obituary text was nowhere to be found on obitmichigan.com. I had to go to the Menominee County Historical Society and request a photocopy of the original newspaper clipping from their microfilm collection. That took about three weeks and cost me forty dollars for shipping and handling.
If you are looking for records before 1921, this site will not help you much. The Michigan Department of Vital Records did not start maintaining centralized death certificates until 1921. Before that, records were kept at the county level or not at all in some cases. There are occasional newspaper obituaries that surface for earlier periods, but they are scattered across different digitization projects and not always indexed properly. I spent a month searching for a death in 1887 and eventually found the obituary by manually scanning through microfilm at the Detroit Public Library. It was a pain, but it was the only way to get the full text. The digital indexes had the name and date, but not the content of the notice itself. Download access is another area where expectations need to be managed. The site does not offer bulk downloads or API access for most of its content. You can print individual pages or save them as PDFs through your browser, but there is no way to export a full county year at once. If you need to build a dataset of several hundred records, you are going to have to do it manually. I wrote a simple Python script that uses Selenium to scrape the county browse pages and export the results to CSV, but that only works for the basic metadata fields. The full obituary text is embedded in an iframe that the scraper cannot reach without simulating a full browser session, which slows things down considerably. It took me about four hours to pull three years of Muskegon County records this way. Doing it by hand would have taken closer to two days. The interface itself is functional but dated. It was built on an older platform and has not received significant updates in years. Search results do not load dynamically. You have to click through each page manually. Filters are limited. There is no way to sort by age or by relationship to surviving family members. You can filter by name, date range, and county, and that is about it. If you are doing genealogical research, you will find yourself constantly refining your searches and going back to adjust parameters. I keep a notepad open next to my monitor and write down every search term I try so I do not repeat myself. It sounds crude, but it saves a lot of time compared to trying to remember what you searched for last week.
Get the Full Details
One counter-intuitive insight about using this site is that the most useful records are often not the ones you are initially looking for. When I was researching one family, I kept focusing on direct line ancestors and ignoring collateral lines. But some of the most valuable information came from obituaries of distant cousins, aunts, and uncles. Those notices often included detailed family tree information that the direct ancestor records did not contain. A cousin's obituary might list all the surviving siblings and their spouses, giving you branching connections that you would otherwise have to piece together from marriage records and census data. I started treating every search as a net rather than a sniper shot, and my results improved significantly after I made that shift. There are also edge cases where the data is simply wrong. I found an obituary where the date of death was listed as 1945 but the age at death was listed as 82, which would put the birth year at 1863. But the census records for that person showed they were living in 1900 at age 30, which would make them born around 1870. The discrepancy was probably due to a transcription error when the original record was digitized. I had to cross-reference with the original microfilm to figure out which number was correct. This happens more often than you would think, especially with records from the 1930s and earlier where handwriting can be difficult to read. The digitization quality varies widely depending on the source institution. Some counties have professional scanners. Others have someone with a handheld device going over faded paper in a poorly lit basement. If obitmichigan.com is not giving you what you need, there are alternatives. The MichiganGenWeb project maintains volunteer-run databases for many counties and often has transcripts that are more detailed than the official records. The Detroit Public Library has a substantial genealogy collection with access to newspapers that are not digitized elsewhere. Local historical societies in the county where your ancestor lived are usually the best resource, but they are also the hardest to access if you do not live in the area. I have mailed requests to six or seven different societies and gotten responses from only three of them. The ones that respond fastest are usually the larger urban societies with dedicated staff. Rural ones tend to rely on volunteers who may not check their email regularly.
The bottom line is that obitmichigan.com is a useful tool but it is far from complete. It covers a significant portion of Michigan deaths from 1921 to the present, with varying depth depending on the county and era. For pre-1921 research, you will need to supplement it with other sources. For post-1950 research, the coverage is generally good but not perfect. The biggest limitations are the lack of bulk export, the dated interface, and the gaps in county participation. If you approach it with realistic expectations and a willingness to dig into the raw data when the search box fails, it can save you a lot of time. If you expect it to be a one-stop solution for all Michigan genealogical research, you will be disappointed within the first hour. I still use it regularly. I just do not trust it blindly anymore. Every date and name you pull from the system should be verified against at least one other source before you lock it into your family tree. The convenience of having these records online is real, but the convenience comes with a cost in accuracy that you have to manage yourself. There is no quality control layer between the original record and the digital display. You are the final check, whether you want to be or not.
