Your phone has 47 contacts from last week's conference. Six of them matter.

47 captures. Six follow-ups. The pile is not the problem.

You worked the room. Your phone, by Friday afternoon, has 47 contacts you didn't have on Monday. Some are people you spoke with for two minutes at a coffee station. Some are people who asked the question that told you they're already shopping. By Sunday night, you cannot tell which is which.

The usual response is to push through it. Send the same friendly note to all 47 and let the universe sort out who replies. This is the wrong response. The forty-one who got the generic note remember you as the person who sent the generic note. The six who had a live problem remember you as one more capture-and-blast operator. The pile didn't get smaller; the brand got worse.

The pile is correct. Conferences produce piles. What's missing is a triage step between the pile and the keyboard. Four questions, applied honestly, separate the live deals from the polite handshakes. The answers map onto a 5-day, 14-day, 30-day follow-up schedule. The forty-one get a single line and a clean exit. The six get the version of you that took the meeting seriously.

Why most contact apps are useless for triage

The category leaders ship the same product shape. Capture screen takes a card. Card-detail screen stores the contact. Contact-list screen alphabetizes the result. None of those screens models relevance. The app treats the person you met at 9:14am and the person you met at 4:47pm as the same kind of row in the same kind of list.

This is the same pattern the OCR-vs-action piece names: the scan that goes nowhere, the capture screen disconnected from a follow-up screen, the contact list that confuses storage with relationship. The four-question triage is the operator's workaround for the category's failure to model context. You apply the questions because the app won't apply them for you.

A contact app worth using gives you somewhere to put the answers. Tags for question one. A field for the commitment in question two. A timing column for question three. A flag for question four. Most existing scanners have no place for the answers.

Question 1: Did they ask a specific question, or a generic one?

*What did they actually ask you?* This signal happens at the start of the conversation and decays first.

A specific question signals real intent. "How does your event-mode capture work when the venue's wifi is dead?" is a specific question. The person asking has a problem and has thought about it long enough to phrase the failure mode. They are not making conversation; they are interviewing your product.

A generic question is polite interest. "So what does your app do?" is the social move that comes before pulling out a phone. It does not predict a follow-up that lands.

The action: tag the contact with the question, in the smallest number of words that carry the meaning. "Wifi-dead capture" beats "asked about offline mode". The tag is what your future self uses, three days later, to remember why this person mattered. If the only tag you can write is "asked what we do", the contact belongs in the bottom tier.

Question 2: Did you commit to anything concrete, or did you both say 'let's stay in touch'?

The second question audits the end of the conversation, not the beginning. *What did you actually agree to do?*

A concrete commitment is a deliverable with a verb in it. "I'll send you the integration spec by Tuesday." "You'll forward me your buyer's name and I'll reach out next week." Concrete commitments survive the noise of the conference because both sides walked away with an item on a list.

"Let's stay in touch" is not a commitment. Politeness sounds like a commitment in the moment and is invisible 72 hours later. If you cannot write down what you owe the person, you owe them nothing, and they don't expect anything from you.

The action: write the commitment in the contact's note field before you leave the venue, or before the end of the day at the latest. The verb-noun pair is what makes the follow-up obvious. Without it, the follow-up has to invent a reason to exist, which reads as cold outreach.

Question 3: Will the context they're in change in the next 30 days?

The third question is about timing, not warmth. *Is something specific about to happen that makes this person reachable now and not in three months?*

A changing context is a forcing function. New role starting in two weeks. Fundraise closing this month. A Q3 initiative that needs to be staffed before the quarter ends. Board presentation in six weeks. Each is a window that closes. The follow-up has to land while the window is open, or it lands as background noise after the decision was made.

A static context is the opposite. They're happy in their setup. Between projects. The contact is real; the timing is not urgent. Static contexts belong in long-term nurture, not in the 5-day window.

The action: tag the contact with the timeline cue. "Closes round in 4 weeks" or "new role June 1" or "static, touch again in October". The dated context determines whether the contact gets a follow-up this week or a quarterly check-in. Most apps do not have a field for this; the workaround is putting the date in the note and the keyword in the tag.

Question 4: Can you do something useful for them in the next 5 days?

The fourth question is the operator's question. It separates a follow-up from a check-in. *Do you have something specific to send, introduce, share, or do, in the next five days, that benefits them and not just you?*

If yes: the follow-up writes itself. "You asked who I'd talk to about the wifi-dead question; I'd start with this person, copying you." The follow-up is a delivery, not a request for a meeting. Reply rates are high because the recipient gets value before they're asked for time.

If no: it's a generic touch, treat it as one. A one-line note that does not request a meeting and does not propose next steps. The contact moves to the long-term-nurture queue.

The action: be honest about which kind of follow-up you can write today. The pile of 47 turns into six who get a real follow-up, twelve who get a one-line touch, and twenty-nine who get a clean exit. The clean exit is a tier; the contact stays in your records, the conversation does not get forced.

The 5-day, 14-day, 30-day decay schedule

The four questions sort the pile. The decay schedule paces the follow-ups against the 72-hour memory cliff that runs in both directions: their memory of you, and your memory of them.

Days 1 to 5. The six who scored highest (specific question + concrete commitment + closing timing window + deliverable in hand) get their follow-ups inside this window. After day five, you are competing against the next week's noise.

Days 6 to 14. Middle tier (a strong yes on two of the four questions, plus reachability). A check-in note with a specific reference to the conversation, plus an open-ended next step. Reply rates drop by half versus the 5-day window.

Days 15 to 30. Long-tail nurture. Static contexts and polite interest. A short note referencing the meeting and a useful link. Set a 90-day reminder for the next touch.

After day 30. Anything that did not get sent inside the 30-day window does not need to be sent. Reaching out at day 45 with "hey, just following up from SOF Week" reads as the cold outreach it is. The honest move is a fresh hello in three months, on a new pretext.

What this looks like as a workflow

The four questions and the decay schedule turn into a workflow only if the contact app has somewhere to put the answers. In Met, the four questions map directly onto the contact-detail surface: tags carry the question-1 signal and the question-3 timing cue; the note field holds the question-2 commitment and the question-4 deliverable; the follow-up draft view turns the answers into a sent message instead of a stored row.

The pattern works in any app that lets you tag freely, write a note that survives the capture, and surface a draft as the screen after the scan. Met starts at the follow-up. The draft is generated from the tags, the note, and the meeting context, in the tone the reader picked. The contact list exists, but it isn't the destination. The destination is the message that gets sent inside the 5-day window.

Pick six. Permission to delete the rest.

The hardest part of the four-question triage is not the answering. It's the giving-yourself-permission part. Forty-one of the 47 captures from last week do not warrant a follow-up. That is the correct outcome. The work is selecting honestly, not engaging exhaustively.

Pick six. Send the deliverables in the 5-day window. Send the soft touches in the 14-day window. Schedule 90-day check-ins for the long-nurture tier. Drop the rest from the active queue (they're still in the records; they're just not in the follow-up rotation). By the time the next conference rolls around, the six from this one have either become customers, become introductions to customers, or become clear nos. That outcome is worth more than 41 generic notes that produced 41 polite non-replies.

The pile is not a measure of how well the conference went. The six are. Triage hard. Follow up specifically.