Call logging: what to record so the next call is easy
Date, time, outcome, next action, what was promised. Why spoke to customer is not a log entry, and how to write a note a colleague can use weeks later.
· 5 min read
The cost of a log nobody can use
Most call logs are kept and most are useless, which is a worse outcome than not keeping one, because the effort was spent and the benefit never arrived.
The symptom is recognisable. A follow-up call is due, the log is consulted, and it says something like spoke to customer, will revert. That entry records that a call happened and nothing about what was said, so the caller has two options: start the conversation over, which annoys the person and wastes the earlier call, or improvise around a gap and hope nothing important is missed.
The underlying problem is that the note was written for the person writing it, in the moments after a call when everything was still in mind. Written that way, a log is a memory aid that stops working within about a week.
A log is worth keeping only if it is written for someone who was not on the call — which includes you in three weeks. That single shift in audience is most of what separates a useful log from a ritual.
The six fields that do the work
A log entry needs six things, and their order matters less than their presence.
Date and time, because the interval since the last contact determines what the next call should be.
Who was actually spoken to, which is not always the person intended — a call answered by a colleague is a different event from a conversation with the decision-maker, and conflating them is how a sequence gets misjudged.
Duration, roughly. A two-minute call and a twenty-minute call are different kinds of contact, and this is the cheapest available signal of engagement.
Outcome, in specific terms rather than as a status.
Next action, with a date, and whose action it is. This is the field that most often decides whether anything happens at all.
What was promised, by either side, recorded exactly — a price, a delivery window, a document to be sent, a discount mentioned in passing.
The last one is the field most often missing and the most expensive to lose, because an unrecorded promise is still a promise the customer remembers.
Why spoke to customer is not an entry
It is worth being concrete about what makes a note useless, because the failing entries all share a structure: they record the caller's impression instead of the content.
Spoke to customer records that a call occurred. Went well is an assessment with no evidence attached, and it is frequently wrong in hindsight. Will get back to us omits when, and about what. Not interested closes a record without saying what they were not interested in, which matters because the objection is the useful part.
The replacement is not longer notes. It is notes containing specifics: what the person said in their own words about their situation, the actual objection, the actual date they mentioned, the actual figure discussed.
A practical test that catches most bad entries: could someone else make the next call using only this note? If they would have to ask you what happened, the note has not done its job. Applying that test at the moment of writing costs a few seconds and changes what gets written.
Writing for the colleague who picks it up in three weeks
The audience for a log entry is a person with no memory of the call, possibly not the person who made it, reading under time pressure just before dialling.
That implies several things about how to write. Avoid private shorthand and abbreviations that made sense at the time. Avoid pronouns without antecedents — he said he would check with him is a common and unusable construction. Name people and roles explicitly, since a name means nothing to a colleague without the role attached.
Quote rather than summarise where the exact words matter, particularly for objections and commitments. A summary of an objection is already an interpretation, and interpretations lose the detail that would have let the next caller respond to what was actually said.
And separate fact from inference, explicitly. Said the budget is decided in the next quarter is a fact. Seems keen is an inference, and it is worth recording as such rather than as though it were reported. A log that mixes the two produces a caller confidently referring to something the customer never said.
Writing it immediately, and keeping it short enough to sustain
The two practical constraints on logging are that memory decays fast and that any process taking too long stops happening.
Write the note immediately after the call, before the next one. The decay is steeper than people expect: specifics, names, exact figures and the phrasing of an objection are largely gone within an hour, and what remains is a general impression, which is the least useful part. A note written at the end of the day is a note about feelings.
Keep it short enough to survive a busy period. Six fields and a few sentences is about the limit of what gets done reliably when there are fifteen calls to make. An elaborate template with a dozen fields produces complete records for a fortnight and then empty ones.
One useful trick: log the next action first, while the call is still ending. It is the field with the most consequence and the one most likely to be lost, and writing it first means an interrupted note still contains the important part.
Where the note lives matters less than that there is one place for it.
What a log makes possible beyond the next call
A log kept properly answers questions no individual call can, and this is the return that justifies the discipline.
Across a few dozen entries, patterns become visible: which objection actually arrives most often, how many contacts a typical decision takes, which sources of contacts produce conversations and which produce nothing, whether calls at particular times reach people. Each of those changes a decision, and none is knowable from memory, which systematically over-weights recent and dramatic calls.
It also protects the business. A record of what was promised, by whom and when, resolves the disagreements that otherwise turn on two people's recollections — and it protects the customer equally, which matters if a commitment was made that the business needs to honour.
And it makes the work handoverable. A business whose call history exists only in one person's head cannot add a second caller, cover an absence, or survive that person leaving.
Automix's Wani, a planned voice follow-up agent, is documented at /docs/wani. It is not available to customers today, so the log described here is kept by whoever makes the call.
Common questions
What is the minimum I should record for each call?
Date and time, who you actually spoke to, the outcome in specific terms, the next action with a date and an owner, and anything promised by either side. Rough duration is worth adding because it is the cheapest signal of how engaged the conversation was. Fewer fields than that and the next call starts from nothing.
Why is what was promised its own field?
Because it is the item with consequences and the one most often lost. A price mentioned in passing, a delivery window, a document you said you would send — the customer remembers all of these whether or not you recorded them. An unrecorded promise is still a promise, and disputes about them turn on whoever has a record.
When should I write the note?
Immediately after the call and before the next one. Specifics, exact figures and the phrasing of an objection are largely gone within an hour, leaving a general impression that is the least useful part of the call. If you must be brief, write the next action first, since that is the field with the most consequence.
Does a spreadsheet work, or do I need proper software?
A spreadsheet works, and so does a notebook, provided there is exactly one place where call notes live and everyone uses it. The common failure is not the tool but notes scattered across a phone, a diary and someone's memory, at which point no history can be reconstructed regardless of how good any individual note was.
Related pages