Measuring travel client retention in a small agency
Measure retention with defined clients, cohorts and repeat events, using a practical worksheet that separates booking share from actual returning households.

Measure retention with a defined client unit, starting cohort, repeat event and observation window. Keep that measure separate from the proportion of this month’s bookings made by past clients. Travel purchases are not necessarily annual, so a quiet client should not automatically be labelled lost without considering the service and timing being measured.
Measure retention from your agency's actual client and transaction records, then use Nowah to improve the planning service those measures describe. Compare a returning client's current route options and explain the tradeoffs clearly. That gives a retention review a practical next experiment without confusing a planning conversation with evidence that another booking occurred.
Define the client and the repeat event
Choose a consistent unit such as an individual or household under your actual process. Several travellers in one household trip are not automatically several independently retained clients. One person making several bookings can also distort a simple booking count.
Specify the event: another paid consultation, a subsequent confirmed trip or another clearly defined service. Record cancellations and unresolved cases according to that definition rather than switching between enquiries and bookings to improve the result.
Keep the source for each event traceable to the actual agency record. A client saying they may travel again is an interest stage, not a confirmed repeat purchase. A supplier booking belonging to the same trip may be another component rather than another journey. The measure should describe what the agency actually wants to understand, with enough detail that another person can reproduce the count.
Choose a cohort and a fair observation window
Group clients by a defined starting event or period, then follow the same group to a stated cutoff. For example, households with a first completed trip in a specified year can be observed for a later repeat event.
Keep the window visible. A client first travelling late in a period may have had less opportunity to return than someone who travelled early. Do not compare unequal observation periods as though they represent identical behaviour. If you use a local method to handle that difference, document it rather than implying a universal benchmark.
Separate no repeat event yet from an explicit decision not to use the agency. A client can remain interested, have another planning cycle or simply not need travel now. Contact preferences and privacy also matter. Do not turn a measurement gap into permission to pursue someone repeatedly or treat silence as a reason to add them to more messages.
Keep related metrics distinct
Cohort repeat proportion asks how many defined starting clients have reached the repeat event by the cutoff. Repeat-booking share asks how many bookings in a period came from prior clients. Distinct returning-client count asks how many actual client units those bookings represent.
All can be useful, but they answer different questions and use different denominators. Do not label the same-period ratio of repeat bookings to all bookings as the retention rate of last year’s clients unless that is genuinely the defined cohort method.
Keep revenue, contribution and service quality separate too. A high repeat share may arise because new enquiries fell, not because more clients returned. A returning client may make several small purchases, while another makes one substantial trip. Use the actual data and stated question rather than an isolated percentage presented as a complete picture of loyalty or profitability.
Worked hypothetical example: two different percentages
A fictional agency defines its client unit as a household. Twenty households completed their first trip in 2025. At a stated 31 December 2026 cutoff, eight have another confirmed trip that meets the agency’s definition, five have active planning enquiries and seven have no further qualifying event recorded. The categories total twenty.
The observed repeat proportion for that defined cohort is 8÷20=40%. The five active enquiries are not added to the numerator as bookings. The result is qualified by the observation period and method; it does not prove the other twelve households are permanently lost.
In a separate fictional monthly activity view, the agency has thirty confirmed trip cases, nine from past clients. The repeat-booking share is 9÷30=30%. Those nine cases came from six distinct households, so they are not nine independently retained households.
The forty-percent cohort result and thirty-percent booking-share result can both be correct because they answer different questions. Neither is a product benchmark or evidence that a reminder, discount or planning tool caused the outcome.
After completing the two percentage calculations, use Nowah for one returning household's newly confirmed trip brief. Compare two route directions and record what the client asks to change using the preference update questionnaire. The figures remain fictional counting examples; this separate planning task shows what the agency could improve before observing a genuine future outcome.
Copyable retention measurement worksheet
Use this in your normal reporting process:
- Business question and intended decision:
- Defined client unit:
- Starting cohort and source event:
- Qualifying repeat event:
- Observation start and cutoff:
- Treatment of cancellations, duplicates and unresolved cases:
- Cohort denominator and repeat numerator:
- Active-enquiry and no-event counts kept separately:
- Distinct returning clients in any period view:
- Repeat bookings and all bookings in that same period:
- Source records and missing-data limitations:
- Service mix and timing context:
- Action proposed and review date:
Add a calculation check: are numerator and denominator from the same definition and window? Does the label match the actual measure? Could another authorised reviewer reproduce the result from the source records?
Keep private identity and financial information in approved secure records. Reporting usually needs client references and aggregate outcomes, not another unrestricted copy of traveller documents or payment details.
Inspect data quality before drawing a conclusion
Look for duplicate households, renamed records, group enquiries counted as several clients and supplier components counted as separate trips. Use a documented method when the relationship or client unit changes rather than silently merging records to improve a metric.
Preserve cancelled and unresolved stages accurately. A booking later cancelled may need separate reporting under your definition, and a pending case should not disappear from the denominator without explanation. Record missing data rather than replacing it with a convenient assumption.
Reconcile the source events in the client communication history before counting a repeat action. Then bring only the relevant current travel inputs into Nowah for research, keeping reporting and planning distinct. A useful day outline can support the next consultation, but the actual transaction record establishes whether that consultation becomes the defined repeat event.
Use the result to choose a relevant service action
A low observed repeat proportion may prompt a review of service follow-through, contact relevance or the time clients actually take between trips. It does not identify the cause by itself. Inspect actual feedback and case histories with appropriate privacy and scope.
If clients have requested another planning conversation, prepare a current brief rather than repeating the last itinerary automatically. If they asked to pause contact, respect that in the actual workflow. Do not treat every unbooked household as a target for more messages.
Choose one service experiment from the retention review: for example, use Nowah to compare a familiar pace with a route that reduces hotel moves. Use the planning-tool value assessment to define the task you will compare alongside the client outcomes. Measure real replies and subsequent events separately, rather than treating a promising itinerary as a retention improvement already achieved.
Copyable Nowah repeat-brief prompt
Help a travel adviser develop a fictional current brief for two adults considering eight Portugal hotel nights,5–13 November 2027, Lisbon five and Porto three. Origin London with actual flights separate. Ground budget EUR 3,600 excludes international flights and personal spending. The clients have requested a new planning conversation; no historical retention data or past preferences are supplied. List current priorities, rooming, transport, suitability and supplier facts to verify. Do not infer loyalty, repeat-purchase probability, commercial impact, available offers or purchase authority.
Review the output against the new brief. The Nowah features page describes supported planning, editable itineraries, map/timeline views and relevant searches. Use those capabilities for current service delivery while the agency’s actual records establish retention measures.
Review trends with the method unchanged
Compare like definitions and observation windows, noting changes in service mix, capacity or data quality. Small numbers can move substantially from one case, so show counts as well as percentages. Do not present a local result as an industry standard.
Keep the reporting action specific and testable, such as improving agreed follow-up or clarifying preference updates. Review service obligations and client choices alongside commercial outcomes.
Use Nowah for planning support and a transparent worksheet for retention measurement. The useful number describes an actual cohort and event, with limits clear enough to guide a proportionate business decision.




