Research ·
Vendor follow-up intervals and ticket age in outsourced queues
How the interval between vendor follow-ups changes the age a customer sees, and how to choose an interval the desk can actually keep.

Key stats
wait components reported
follow-up interval per case class
oldest cases read per review
Methodology and findings
Question and scope
This review asks how the interval between vendor follow-ups changes the age a customer experiences in an outsourced help desk. The population is tickets that wait on a third-party supplier for a fix, a confirmation, or an approval. The unit is one ticket with at least one recorded vendor contact during a declared observation window. The question matters because a ticket can be actively managed while the customer sees no movement for days.
The scope excludes tickets where the vendor is the customer's own team and excludes work with no recorded contact attempt. It also excludes waits that the desk cannot observe, such as a supplier's internal queue. Those boundaries keep the review on contacts the desk controls. A finding should help a reader set an interval and explain it to a client rather than describe a vendor's performance.
What a follow-up interval measures
A follow-up interval is the planned time between one vendor contact and the next. It is not the vendor's response time, and it is not the age of the ticket. The interval describes the desk's own discipline. A short interval can produce more contacts without producing faster fixes, while a long interval can leave a stalled case invisible for a week.
The customer experiences the interval as silence. Even when the desk is working, the absence of an update reads as neglect. That gap between internal activity and customer perception is the center of this review. A measure that counts only vendor replies will miss it entirely, because the desk may be contacting the vendor every day while the customer hears nothing.
Counting contacts honestly
Write down what counts as a contact before extracting data. A sent email, a logged call, a portal comment, and a meeting note are different events with different evidence. Record which channel was used and whether the vendor acknowledged it. An unanswered email is a contact attempt, not a follow-up completed.
Separate the first contact from subsequent ones. The first contact often includes the full case history, while later contacts are short reminders. Mixing them makes the interval look more consistent than it is. Report the first-contact time and the median gap between later attempts as two figures. A reader can then see where the delay sits.
Why a fixed interval fails
A single interval applied to every vendor case ignores difference in consequence. A security fix and a cosmetic question should not share the same cadence. A fixed rule also ignores the vendor's own service window, which may make a Friday follow-up useless. The result is a queue that looks disciplined and a customer who still waits.
A better design sets the interval by consequence and by the vendor's response pattern. High-consequence cases get a shorter interval and a named escalation path. Low-consequence cases can wait longer if the customer was told. The interval should be written into the ticket so the next agent does not have to guess.
The silent wait problem
A ticket waiting on a vendor may show no customer-facing update because the desk has nothing new to say. From the customer's side, that is indistinguishable from being forgotten. A short holding message at each interval costs little and prevents most complaints about silence. The message should state what was asked, when the next check falls, and what the customer can do meanwhile.
These holding messages also create a record. If the vendor later claims it was never contacted, the messages and their timestamps support the desk. If the customer escalates, the sequence shows a managed wait rather than an abandoned one. The communication is evidence as well as service, and it should be stored with the ticket rather than in a personal mailbox.
Reading ticket age with vendor waits
Total ticket age mixes desk time, vendor time, and customer time. That mixture cannot support a fair comparison between agents or between periods. Report the three components separately, even when the tool only stores a single open date. A manual sample can establish the split for a report that the tool cannot produce.
Look at the distribution rather than the average. A small number of very old vendor tickets can lift the mean while most cases close quickly. Report the median, the count above a threshold, and the oldest case. The oldest case usually reveals whether the escalation path works at all.
Where the records break down
Vendor contacts often happen in email threads that never reach the ticket. When that happens, the ticket age looks unexplained and the next agent repeats the contact. Export the relevant thread or summarize it in the ticket at the time of contact. A summary written later is less reliable than one written while the details are fresh.
Some vendors respond through a portal that the desk cannot export. In that case, record the portal reference and the visible status at each check. A screenshot with a timestamp is weak evidence but better than nothing. Note the limitation in any report that relies on it.
Setting an interval you can keep
Choose an interval the desk can sustain on its busiest week, not its quietest. A two-day promise that slips to five days teaches the customer to distrust every later update. A four-day interval that holds is more credible. Test the interval for a month before writing it into a client agreement.
Assign the follow-up to a named queue rather than to an individual. A personal reminder disappears when the agent is absent, while a queue item returns to the team. Set the check time in the same system that holds the ticket. The interval should survive a shift change, a holiday, and a staff departure.
Reporting the wait to the client
Report the number of vendor-dependent tickets, the median interval, the number of missed follow-ups, and the oldest open case. Explain any change in the interval before presenting a trend. A shorter interval that produced more contacts but no faster fixes is still a useful finding, because it shows where the constraint actually sits. Report the same figures for the previous period so the reader can see direction rather than a single snapshot.
Avoid claiming that a new cadence caused faster resolution unless the comparison holds request types and vendor mix constant. Vendor behavior, product changes, and case complexity can move the result without any change in the desk's work. State those limits beside the number. A short note on the comparison limits is part of the report, not an apology for it.
Limits of the review
The review cannot measure what happens inside the vendor's queue. It cannot recover contacts that were never recorded, and it cannot separate a slow vendor from a vague request without reading the cases. The finding describes the documented cadence and its visible effect. A desk should say so rather than imply control it does not have.
Reviewer judgment also affects which cases count as vendor-dependent. Define the boundary with examples and check a sample of classifications. A case that moved between vendor and internal teams may belong in a separate category. Record that decision rule with the definition so a later review can follow it.
Applying the review
Take one month of vendor-dependent tickets and extract each contact attempt with its timestamp. Calculate the gap between attempts and compare it with the stated interval. Read the ten oldest cases and look for missed checks, missing holding messages, and unclear escalation paths. The exercise should produce one change, such as a named queue or a shorter high-consequence cadence.
Apply the change for a month and report the same figures. Compare the interval within the same request type and note any shift in vendor mix. The aim is a wait the customer can see and the desk can defend. A modest, documented cadence is better than a short promise that nobody keeps.
Comparing two vendor queues
Two vendor queues can show the same median interval and behave very differently. One may hold most cases for a few days, while the other holds a small group for weeks. Compare the share above the interval threshold before drawing a conclusion. The share reveals the stalled cases that a median hides.
When a queue has a long tail, read the cases rather than adding more reminders. The cause may be a missing escalation contact, an unclear request, or a vendor that only responds to a different channel. Each cause needs a different fix. A blanket reminder schedule adds work without changing the tail.
Sources
- ISO 18295-1:2017 customer contact centres - Contact-centre performance and service measurement concepts.
- Atlassian incident management handbook - Vendor and stakeholder communication practices during incidents.
- GAO Assessing Data Reliability - Guidance for assessing the reliability of operational data.
- NIST SP 800-53 Rev. 5 security and privacy controls - Audit logging and record-keeping control concepts.