A Practical Guide To Local Obituary Research
Green Bay's local newspapers have been running obituaries since the late 1800s, and the paper records alone contain more biographical information about this city than most people realize. I've spent years going through microfilm and digitized archives to piece together family histories and community stories, and the process is straightforward once you know where to look and what to expect from the records. The Northeast Wisconsin Historical Society holds the bulk of Green Bay obituary records going back to the 1850s, and the Brown Countycourthouse maintains death indexes that cross-reference with newspaper publications. The Green Bay Press-Gazette has an online archive, but it only goes back to 1997 in full-text searchable form. Before that, you need to use their microfilm reading room or work through the county records directly. I started working with these materials about ten years ago when I was researching a client's family connection to the Brown County packing industry. The first thing I learned is that obituary records from before 1940 are incomplete for working-class residents. Wealthy families and established business owners routinely paid for lengthy death notices. Immigrant laborers and seasonal workers often received nothing more than a one-line mention or were skipped entirely. That gap in the record is significant and it skews any narrative about who mattered in Green Bay's history unless you account for it explicitly.
The actual scanning process involves pulling the microfilm for the relevant date range, photographing each obituary page with a DSLR camera on a copy stand, and then running OCR software on the images. The OCR work is where most people hit problems. The Press-Gazette used a gothic typeface for decades, and microfilm degradation makes many characters illegible. I typically run the images through Tesseract with a custom trained model for this newspaper's typography, which improves accuracy from about 62 percent to roughly 89 percent on clean films. Worn or spliced reels drop to around 54 percent even with corrections. Here's a detail beginners consistently miss: cross-referencing the obituary with the Wisconsin Death Index at the State Historical Society of Wisconsin adds middle names, maiden names, and exact dates of death that the newspaper version omits. The newspaper obituary might list "Mary Kowalski, age 72" while the state death certificate shows "Marie Therese Kowalski-Larson, age 72 years, 3 months, 14 days." That maiden name information alone unlocks entire branches of family history that the obituary never mentions. I found three separate cases where a woman's obituary listed her husband's surname as her only identifier, and the state index revealed she was the daughter of a prominent early Green Bay settler. The digital repositories available for this work include the press-gazette.com archive with a paid subscription running about $12 per month, the Wisconsin Historical Society's genealogy portal which requires a paid membership around $40 annually, and FamilySearch which has a growing but incomplete collection of Wisconsin obituaries. FamilySearch's Green Bay coverage currently spans approximately 1902 to 1968 in sporadic completeness, so you should treat it as a supplementary source rather than a primary one.
There is a practical bottleneck with the microfilm approach that nobody talks about. The reading room at the Northeast Wisconsin Historical Society is staffed by volunteers on weekdays, and their hours changed three times between 2019 and 2024 without clear public notification. I drove forty-five minutes there once and found a handwritten sign on the door that said "moving to new location, contact info on website" with no new location listed anywhere. Call ahead before you go. The phone number is listed on their current website and takes about two minutes to confirm whether they're open that day. For people doing this research professionally, investing in a flatbed scanner with a transparency adapter pays for itself within a week compared to the microfilm route. I use a Epson Perfection V850 Pro, and it produces files large enough to recover details from damaged or faded print that microfilm readers blur into unrecognizable noise. A single obituary page scans in about forty-five seconds, and the resulting TIFF file is roughly 200 megabytes. You need adequate storage, but the quality difference is substantial for difficult-to-read copies. Another counter-intuitive finding from my experience: church cemetery records often contain more complete biographical data than the obituaries themselves. St. Francis de Sales Catholic Cemetery and Calvary Cemetery both maintain burial registers that list birth dates, parents' names, and sometimes immigration details. These registers predate the newspaper coverage by several decades in many cases. I pulled Calvary records for a 1923 obituary that mentioned a deceased man named Josef Hendricks, and the cemetery register showed he was born in Posen, Prussia, arrived in Green Bay in 1881, and worked as a cooper before starting his own barrel-making shop on Adams Street. None of that appeared in the newspaper version.
Get the Full Details
The main limitation of scanning obituaries as a research method is that the source material itself is incomplete and biased toward certain demographics. You will find abundant records for German-American and Polish-American families who established themselves before 1930. Mexican-American residents of Green Bay, who have been part of the city since the 1930s cannery seasons, are significantly underrepresented in the press obituaries until the 1970s. Native American records from the Menominee and Oneida communities are similarly sparse in local newspaper coverage. If your research focus includes these populations, you need to supplement with tribal records, parish documents, and community organization archives rather than relying on the press-obituary route alone. Organization matters more than speed. I keep a spreadsheet with columns for the full name as printed, variations in spelling, dates of birth and death, parents' names, spouse information, residence address, employer, and source citation for each record. The source citation format should include the newspaper name, date, section, page number, and the archive or database where the record was accessed. Without that level of detail, you will lose track of where specific information came from, and you'll waste hours re-checking sources that you cannot reliably locate again. The workflow typically takes about four hours for a comprehensive search covering one decade of Green Bay obituaries for a single surname, including scanning, transcription, verification against death indexes, and cross-referencing with cemetery records. Most of that time is spent on verification and cross-referencing, not on the scanning itself. The actual image capture phase for roughly 300 pages of obituaries takes about ninety minutes with the flatbed scanner setup I described.
For those who want to access the Press-Gazette archive directly, the subscription gives you full-text search across their digital backfile. Use Boolean operators to refine results, combining surname with occupation or address fields. A search for "Hendricks AND cooper AND Green Bay" pulls relevant records faster than scanning through chronological listings page by page. The search function indexes from 1997 onward reliably, and for earlier years it has a variable success rate depending on how thoroughly the microfilm was digitized for that particular time period. The most useful recent development for this type of research is the Brown County Genealogical Society's ongoing transcription project, which has made thousands of pre-1950 obituaries searchable by name through their website at no cost. Their coverage is not complete, but it includes a disproportionate number of the earlier German-language obituaries that appear in bilingual newspapers and are harder to find elsewhere. I've used their database as a starting point for about half of my current research cases because it surfaces names that the Press-Gazette's own search tool misses. If you're doing this work for genealogical or historical purposes, expect the records to be fragmentary and work your findings against multiple source types rather than treating any single obituary as authoritative. A single newspaper death notice rarely contains enough information to establish facts definitively. The death certificate, cemetery register, census records, and any published biographical sketches should all be checked before drawing conclusions about the person's life, family connections, or community role. That multi-source verification is what separates reliable local history from speculation built on incomplete obituary fragments.
