Small-business outsourcing research

How old are open customer promises before a follow-up date is set?

A protocol for measuring open customer promises, the time before a follow-up date is recorded, and the promises that quietly expire.

How old are open customer promises before a follow-up date is set?

Research question and scope

This protocol asks how old an open customer promise is before a follow-up date is set. The unit is one promise made to a customer, identified by the customer, the request, and the date the promise was made. The study measures whether a follow-up date exists, how long it took to record, and whether the promise was kept. It does not evaluate whether the promise was generous or whether the customer was satisfied.

Set the scope to one small business, one support queue, and one promise template. Record the channels, the business hours, and the people who may make a promise at the start. If the template changes during the window, keep the earlier definition for the full sample or analyze the change as a separate period. A new reminder can look like a better follow-up when it is only a new alert.

Population and unit of analysis

Define the population as every promise made during the window, including promises made by the owner and promises made by an assistant. The unit is the promise, not the conversation, because one conversation can contain several promises. A promise that is restated with a new date is the same unit with a revision. A new promise about a different request is a new unit.

Record the channel, the request type, the customer type, and whether the promise was made in writing during the conversation. These strata matter because a delivery date and a callback time carry different risks. Include promises that were kept early, because the base rate is part of the finding. Test the unit definition on two ordinary promises before collection begins.

A promise can be made to a customer and repeated to a colleague without ever reaching the queue. Record the source of the promise and the person who was told, not only the customer-facing message. A promise that exists only in a meeting note is still a commitment. The unit should include it when the note names the customer and the request.

Definitions that make the measure reproducible

Define a promise as a statement that the business will do something for the customer by a stated or implied time. Define a follow-up date as a calendar date in the queue record, not a relative phrase such as next week in the message. Define an open promise as one with no recorded completion or closure. Define a late follow-up as a first contact after the recorded date.

Do not count a general statement such as we will look into it as a dated promise unless a date is recorded. Do not count an internal reminder as a customer follow-up. A promise that is withdrawn with the customer's agreement is closed, not late. Each definition should include one positive example and one boundary example.

Evidence standard and data dictionary

For each promise, capture the customer, the request, the channel, the promise date, the follow-up date, and the owner. Add the first contact after the promise, the outcome, and whether the customer confirmed the result. Record whether the promise was made in writing and where the message is stored. Keep the message with the queue record.

Also capture whether the follow-up date was set at the same time as the promise or added later. A date added after the customer asks is a weak control. If the promise changed, record the old date and the reason for the change. A second reviewer should be able to reconstruct the timeline from the record.

A follow-up date can be set by the assistant, the owner, or an automation, and the source changes the meaning of the promise. Record who set the date and whether the customer agreed to it in writing or on a call. A date the customer never heard is not a customer promise, even when the queue treats it as one. The data dictionary should state the difference.

Sampling and collection

Draw the sample from the support queue rather than from a list of complaints. Record the export date, the filters, the row count, and the selection rule. Include promises from every channel and request type so the result is not dominated by one source. If the population is ordered by promise date, a fixed interval spreads the sample across the window.

Collect the promise as it existed when it was made, not as it looks after a later update. If the follow-up date was added days later, record the addition time and analyze the original gap separately. Do not infer a kept promise from a closed ticket. Note the collection date separately from the promise date.

Analysis and reporting

Report the number of promises, the share with a follow-up date at the time of the promise, the share with a date added later, and the share with no date. Show the counts by channel and request type. Present the median age of an open promise and the share kept by the recorded date. Use the same denominator for every rate.

Inspect the promises with no date and describe the pattern without naming the customer or the assistant. A cluster in one channel may point to a missing field, while a cluster in one request type may point to an unclear process. Compare the result with the promise age, because an old open promise may be a sign of a forgotten handoff. Report the revised promises separately.

A useful additional view is the share of promises that were kept but never confirmed with the customer. A task can be complete while the customer still waits for a message. Show the confirmation rate beside the completion rate and describe the difference. That view helps the owner decide whether the problem is the work or the communication.

A useful cut is the share of promises that were kept but never confirmed with the customer. A task can be complete while the customer still waits for a message. Show the confirmation rate beside the completion rate and describe the difference. That view helps the owner decide whether the problem is the work or the communication.

Limitations and sources of error

The record cannot show whether the customer accepted the promise or whether the business could have kept an earlier date. It can show only what was recorded and when. A kept promise is not proof of good service, and a late one is not proof of neglect. The protocol also cannot capture a promise made in a meeting or on a call that left no note.

Support systems can merge threads, change timestamps, and close tickets without a customer confirmation. A promise that looks closed may still be open in the customer's view. Record those cases and their treatment instead of dropping them silently. If the person who made the promise also confirms it, note the independence limitation and analyze that subgroup separately.

Review cadence and next action

Give the result to the owner and the support lead and ask them to choose one observable change. The change might be a required follow-up date, a daily list of open promises, or a customer confirmation step. Record the owner, the effective date, and the measure that will show whether the change worked. Avoid changing the template and the staffing at the same time.

Repeat the protocol with the same definitions after the change and compare the same fields. If dated promises rose but kept promises did not, the team may be setting dates it cannot meet. If kept promises rose but confirmation stayed low, the customer may still feel forgotten. Close the review when the owner accepts the finding and the next sample is scheduled.

Give the support lead the same protocol and a sample of promises from each channel, including phone notes. A template that works for email may fail when the promise is made by phone or in a short message. Record the review date and the change that was tested. The next sample should use the same fields so the comparison is fair.

Sources

Frequently asked questions

Does a favorable sample prove the routine caused the result?

No. The design is observational and supports a bounded operating decision, not a universal causal claim.

What should an owner review first?

Start with high-consequence defects, missing denominators, exceptions, and cases that required the owner to reconstruct the evidence.

Related research