A notebook can hold a wonderful conversation and still hide it from its owner. I remember the subject, the setting, perhaps even the weather, but not which file contains the detail I need. For a business feature writer, that is a practical problem. Good material becomes harder to check when finding it depends on memory.
My starting point would be a small interview index. One row represents one conversation. The columns record the date, the speaker’s name, their role, the business, the main subjects and the location of the original notes. The index is a finding aid. It should point back to evidence rather than pretend to replace it.
What belongs in the subject column? Plain words that I would actually search for: rent, delivery, training, winter trade, opening hours. A fixed list helps, but it needs room for an unexpected subject. I would rather add a useful term than force every interview into categories designed before the reporting began.
A local script could scan filenames and build the first version of this table. Another short routine could flag missing dates or two records that appear to describe the same interview. Neither task needs a complicated application. A simple spreadsheet, a folder of text files and a few clear naming rules can do much of the work.
Search results still need judgment. If a shopkeeper discussed rising costs last spring, that does not establish today’s position. I would include a follow-up date and a status field such as needs checking, checked or unsuitable for publication. That small distinction prevents an old observation from quietly becoming a current claim.
I would also keep quotation text out of the index unless it serves a clear purpose. A short summary is usually enough to locate the source. Full notes, recordings and contact details belong in their appropriate folders, with access limited to the people who need them. A useful tool should not scatter sensitive material across unnecessary copies.
The best test is an ordinary reporting question. Can I find the three interviews about deliveries, open the original notes and see what still needs checking? If the answer is yes, the system is helping. If every search requires repairing a formula, I have built more machinery than the job deserves.
Once a month, I would review a handful of records against their source files. Broken paths, inconsistent tags and unexplained abbreviations are easier to fix while the conversations remain familiar. The habit matters more than the software. An index earns its place when it gives a writer more time to listen, verify and write clearly.
I would begin with ten interviews, not the whole archive. That smaller sample would expose awkward tags and missing information before they spread. A colleague could try the same search and describe where the labels confused them. Once the routine feels clear, adding older records becomes a manageable piece of editorial housekeeping rather than an ambitious project that never reaches the desk where it is needed every working day.
Top comments (0)