A client mentioned the real pain point once, and the only copy is a two-minute memo you have not opened. You slide your mouse over the files folder right now, no tab shuffling, no prepping a fancy note first. The desk has a half-empty mug of tea and a single scratch pad next to your keyboard, so you don’t need to grab anything extra. You already turned off your Slack notifications 90 seconds ago, so no pop-ups will yank your focus mid-listen.
First Pass: Capture every unfiltered line without edits
Hit play on the memo once, no pausing, just listen all the way through to get the general flow of the conversation. You might notice you forgot half the things said, even if you recorded it just three days ago, that’s normal. Play the memo a second time, and type every single word you hear exactly as it’s spoken, no trimming, no rephrasing, no deleting filler words or half-finished thoughts. If you catch yourself editing a line to sound more polished, stop, backtrack 2 seconds, and type the line exactly as it was said. Include the pauses you hear, the offhand laughs, the parts where the client trails off and says “never mind”, even the mumbled follow-up question you asked in the background that you completely forgot about. If you miss a word, rewind just far enough to catch it, then keep going. This pass is purely about getting all the content out of the audio and onto your screen, no judgment, no curation. For a two-minute memo, this takes no more than 3 minutes total. Illustrative example: if the memo includes the line “uh, I don’t know, we just spend three hours a week chasing people for expense reports and no one even gets why we ask for half the fields”, you type that exactly, no cutting the hesitation or hedging first.
Second Pass: Flag fragments that hold unspoken weight
Now scan the raw transcript you just typed, line by line. You’re looking for lines that make you pause, lines that carry frustration, relief, offhand mentions of workarounds, things the speaker says they “never get around to” or “always have to fight for”, or comments that feel like throwaway lines but hint at a bigger unmet need. You don’t rewrite anything here, just drop an asterisk next to each of these fragments to flag them for the next pass. You use this table to track your work as you go, no fancy formatting required:
| Memo timestamp | First pass raw line snippet | Second pass flag reason | Third pass refined insight |
|---|---|---|---|
| [fill in date/time here] | |||
| [fill in date/time here] | |||
| [fill in date/time here] |
Start by filling in the timestamp for the memo you’re processing now, then paste each flagged raw line into the second column of the table. Add a 3 to 5 word reason for the flag in the third column, no overthinking: “frustration re: expenses”, “offhand workaround mention”, “forgotten unmet need” all work perfectly. If you go through the entire transcript and don’t find any flagged lines, that’s fine. You can close the transcript and move on, no need to force insight where there isn’t any. Most two-minute client memos have 1 to 2 flagged lines, so this pass takes no more than 2 minutes total per memo. Illustrative example: you flag that expense report line, and write “3 hrs weekly unplanned work” as your flag reason.
Third Pass: Reframe flagged lines into actionable, shareable insights
Now go back to each flagged line and its corresponding flag reason, and rewrite it into a clear, concise line that anyone can understand without listening to the original memo or having extra context about the conversation. You don’t add extra assumptions, you don’t extrapolate beyond what the speaker actually said, you just trim the filler and frame the core observation clearly. For the expense report example, the refined insight might be “Client operations team spends 3 unplanned hours per week chasing expense report submissions, and submitters do not understand the purpose of half the required fields”. Drop that refined line into the fourth column of the table for that memo. If a flagged line is too vague to refine right now, note that in the fourth column, and add a short reminder of who to follow up with to get clarity, like “follow up with Sam re: tool access request”. This pass takes roughly 1 minute per flagged line, so for the average two-minute memo, you’ll be done with all three passes in under 8 minutes total, no extra work required. If you get stuck on how to reframe a line to pull out the core observation, you can cross-reference the approach for highlighting unspoken user needs from Find Core Insight if you have it saved, but you don’t need extra resources to finish this right now.
Log the output where you will actually use it later
Don’t leave your completed table row sitting in a random scratch file where you’ll never find it again. Copy the refined insight over to the exact place you’ll need it next: if it’s a client pain point, paste it into the client’s account notes, if it’s product feedback, paste it into the shared feedback repository your team uses, if it’s a personal idea you recorded on a walk, paste it into the folder where you keep project ideas for your side work. You don’t need to file it into 17 different folders or tag it with 12 different labels, just put it in one place you know you’ll look when you’re working on that specific project or client next. If you have a backlog of 10 or more voice memos to process today, you can run all of them through the three passes first, then log all the insights at once to cut down on context switching, but for this single two-minute memo, you can log it right now in 10 seconds. Delete the raw transcript if you don’t need it for compliance or legal purposes, or save it to a dedicated voice memo transcript folder if you do, no need to hoard it in your downloads folder cluttering up your screen. If you only have 20 minutes free right now, set a timer and knock out as many memos from your backlog as you can in that window, and stop when the timer goes off, no guilt if you don’t finish the whole backlog in one go.
Prep 10 seconds before your next voice memo review session
The next time you sit down to process a batch of voice memos, you don’t need to set up a complicated workflow or download new tools. Before you hit play on the first memo, open a blank version of the table you used today, and have the folder or tool where you log final insights open in a separate tab so you don’t have to go searching for it mid-session. If you usually record voice memos while you’re out running errands or walking between meetings, add a 2-second label to each memo right after you record it, like “client X pain point” or “blog post idea”, so you don’t have to guess what’s in each file before you start processing. You can batch this review once a week, or process each memo the same day you record it, whichever fits into your existing schedule without adding extra work. If you only have 5 minutes free, you can process 2 to 3 short memos in that window, no need to block off a full hour to get value from the work. If you come across a memo longer than 10 minutes, split it into 2-minute chunks before you start the first pass, so you don’t get overwhelmed halfway through transcribing. When you finish processing each batch, close all the scratch files you used, so your desktop doesn’t get cluttered with half-finished transcripts you don’t need anymore. You can adjust the number of memos you process per session to fit the time you have available, no minimum requirement to make the work worth your time.