Client searches cleaned
29
News Search Cleanup · Research Brief
News Search Cleanup · What a name search returns
49% of Lexis search results were duplicates or the wrong person
One name search downloads as a single Word file of 500 to 3,000 articles. Much of it is the same story, republished by eight different newspapers. Much of the rest is a different human being who happens to share the name. Clearing that out by hand takes a staffer a week. We built software that does the work. It has cleaned 29 client searches, 33,771 articles in all, and attaches a written reason to every removal.
Articles reviewed
33,771
Duplicate or wrong person
49%
Reached the research file
51%
Figures are from client searches in the 2026 cycle. Candidates are identified only by office and state.
Duplicates and wrong-person articles
Two kinds of junk, and someone has to clear both by hand
A name search returns every article that contains the name. Duplicates make the researcher read the same article again. Wrong-person articles give the researcher facts about a stranger.
The same story, again
31% are duplicates
The same story, in three to eight newspapers
Across the fourteen searches where the duplicate pass was logged article by article, 7,457 of 23,792 were copies of another article in the same file. In one Iowa House race it was 46%, nearly half of what came back.
The wrong person
231 of 481 articles
Other people with the same name
In one Florida House race, 231 of the 481 articles the search returned were about a different human being. They included honor‑roll students, high‑school track athletes, two student newspaper reporters, and the lead singer of an Irish rock band.
How much of it is the same story
About a third of every search is the same story, over and over
One story from a national news service gets republished three to eight times, each time under a different newspaper’s name. A newspaper chain runs a single piece across dozens of its own papers over several weeks. The table shows duplicates in fourteen client searches.
| Race | Articles returned | Duplicates removed | Share |
|---|---|---|---|
| Iowa House seat | 638 | 293 | 46% |
| Arizona governor | 1,999 | 844 | 42% |
| Florida Senate | 2,000 | 840 | 42% |
| Colorado House seat | 2,997 | 1,176 | 39% |
| Arizona House seat | 2,071 | 776 | 37% |
| Pennsylvania House seat | 2,051 | 756 | 37% |
| Second Iowa House seat | 2,049 | 718 | 35% |
| Second Pennsylvania House seat | 2,500 | 784 | 31% |
| Second Colorado House seat | 1,244 | 369 | 30% |
| Ohio House seat | 1,813 | 538 | 30% |
| Nevada governor | 840 | 222 | 26% |
| Third Iowa House seat | 112 | 24 | 21% |
| Second Florida House seat | 481 | 56 | 12% |
| Florida House seat | 2,997 | 61 | 2% |
| These fourteen searches | 23,792 | 7,457 | 31% |
Duplicates were between a quarter and a half of most searches. In the three searches at the bottom, most of the file had already been removed as articles about other people, so very little was left to duplicate. In the second Iowa race, 245 original stories accounted for 985 articles, four copies apiece on average. The most-copied story had far more.
Wrong-person articles
One Florida House race returned 2,997 articles. 115 made the research file.
The candidate’s surname is also an ordinary English word for a job, and it is shared by several accomplished people. A search cannot tell the difference. Of 2,997 articles: 611 were about a different real person, 766 used the surname as an occupation, and 1,434 mentioned the name without saying anything about him. Another 71 were export cover sheets, foreign-language reprints, or copies of an article already in the file.
Under 4% of the search was about the candidate
2,997articles returned
→
115reached the research file
The other 2,882 articles are kept, each with a note naming who it was about, and the client can read that list.
How much is somebody else
Articles about somebody else ranged from 1% to 94% of a search
Duplicates were about a third of a typical search. The share of articles about the wrong person depended on how common the candidate’s name was, and almost no searches fell in the middle. Eleven of the 29 searches carry a categorized exclusion log that names, for every article removed, the person it turned out to be about.
| Race | Articles returned | Not this person | Share |
|---|---|---|---|
| Florida House seat | 2,997 | 2,821 | 94% |
| Third Iowa House seat | 112 | 68 | 61% |
| Second Florida House seat | 481 | 271 | 56% |
| Arizona governor | 1,999 | 83 | 4% |
| Arizona House seat | 2,071 | 85 | 4% |
| Pennsylvania House seat | 2,051 | 57 | 3% |
| Nevada governor | 840 | 21 | 3% |
| Colorado House seat | 2,997 | 41 | 1% |
| Second Colorado House seat | 1,244 | 18 | 1% |
| Second Pennsylvania House seat | 2,500 | 35 | 1% |
| Florida Senate | 2,000 | 27 | 1% |
| These eleven searches | 19,292 | 3,527 | 18% |
9,435 of the 19,292 articles (49%) came out in the two removals together. 5,908 were copies of an article already in the file. 3,527 were about somebody else or never said anything about the subject. The remaining 9,857 articles are the file the researcher reads.
Put it to work
You already paid for the articles. Get the research file.
We run this on client searches today. The News Article Cleaner in the Civly platform takes your Lexis Word export and gives you back the articles with wire-service and chain reprints collapsed to one. It does not yet screen for different people with the same name. Every removed article is kept and listed with the reason it was removed, so you can always say what is and is not in the public record. If you have a search on a shared drive that your team has not had time to read, that is the one we want to see.
Measured on 29 client LexisNexis searches processed during the 2026 cycle, 33,771 articles parsed in total · the 31% duplicate figure covers the fourteen searches whose duplicate pass was logged article by article (7,457 of 23,792) · the 18% wrong-person share and the 49% combined figure both cover the eleven searches that carry a categorized exclusion log naming the person behind every removal (3,527 and 9,435 of 19,292 respectively) · candidates are identified by office and state only, and every figure comes from the engagement’s removal log.