Skip to content
Interface language: English
Imperium / Signal

The customer stopped ordering because the job was done: when retention measures the wrong goal

Distinguish a completed customer task from a pause or an abandoned project before choosing a retention measure.

Article language: English

Concept illustration: a woman rests her hands on the outer two of three round discs mounted on an axis in a navy and red interior.
In this article

A customer stops ordering because they received what they came for. The report marks them as lost. Their account manager is asked to "win them back" and offers further support. The customer replies that everything works. The company has good news, but its metric still looks unhappy.

Suppose a studio prepares sales documents for an equipment manufacturer. The assignment covers a defined set of materials, consistent names and files prepared for an agreed use. The customer accepts the set, uses it and needs no new work for several months. The studio compares them with a customer who orders every month. That comparison may mix two patterns of need even though both have the same customer status in the system.

Retention, meaning continued active customer relationships, makes sense only against a named relationship and period. In a service that solves a finite task, no follow-up order does not explain whether the first job went well. First establish what the firm promised and what the customer actually completed. Otherwise a long-lasting problem may begin to look like success simply because it generates regular invoices.

What was supposed to end?

Name the service before choosing a measure. The studio may sell a completed set, ongoing updates or access to a team for successive tasks. These are different commitments. If a one-off set was agreed, the absence of another order is not automatically cancellation of recurring support. Recurring support may never have been purchased.

The unit in which a service is sold (in Polish) helps distinguish time, a completed item and access. Here there is another question: how will the customer know that this task no longer requires the provider? That should be known before work starts, not invented while explaining a fall in orders.

In the example, completion may mean handing over the agreed set and confirming that it matches the agreed scope. Using it in sales is a later event. A further sales result depends on other conditions. Do not combine these moments into "customer success" if the studio has confirmed only its own handover. The end of the provider's scope need not be the end of the customer's entire task.

The GOV.UK Service Manual relates completion of a digital transaction to a defined task and endpoint. A submitted application can complete the transaction independently of its later outcome. This is a useful distinction between stages: finishing one action does not automatically confirm the next result.

Use names that fit your own scope: "materials handed over", "accepted against the agreed scope" and "used for the intended activity" need not mean the same thing. Decide which event you can actually confirm. If later use is unknown, record the gap instead of filling it with the customer's polite reply to an email attachment.

Silence does not explain why the customer is silent

In the abstract of their 2010 paper, Fader, Hardie and Shang describe the difficulty of distinguishing a finished relationship from a long gap between transactions where there is no formal cancellation. They applied their model to nonprofit donations. The observation concerns activity; it does not transfer the model's assumptions to the studio or calculate customer value.

At the studio, similar six-month gaps can mean different things. One manufacturer uses the completed materials and has not changed its offer. Another abandoned unfinished work and ordered elsewhere. A third still needs the result but paused its own decision. These histories may look the same in a last-order list. The team needs the reason before choosing its next move.

Start with information already in the case. Was the scope accepted? Are corrections open? Did the customer report that the result could not be used? Did they mention a future change that might create another task? There is no need to call every former customer immediately. First stop labelling people lost when their last recorded message confirms completion.

If a conversation is needed, ask about the task's present state, not loyalty to the company. "Is the set being used as agreed, and is anything still making work difficult?" provides a different basis from "Why did you leave us?" The latter has already decided the answer. The customer may not know that a spreadsheet considers the relationship ended when they simply completed the project.

If the task still exists but the customer can now do it themselves, the question is what work remains to commission. Do not confuse that reason for no orders with the end of the need. The conversation may lead to a narrower scope or confirm that help is not currently needed. See the article on clients taking work in-house (in Polish) for a related perspective.

Do not turn no reply into proof of satisfaction. Separate cases with a known ending from those whose later outcome is unknown. An honest report can leave a gap and decide whether resolving it is needed for a particular offer or quality decision.

Do not confuse a completed need with abandoning a difficult task. A manufacturer may no longer need the documents because it paused a new product line. That differs from successfully using an accepted set. It need not be the studio's fault, but it should not be counted as successful completion merely because someone said "finished". Keep the reason and what was actually delivered.

Aeneia conceptual photography

Do not resolve an unknown outcome by counting messages. Five replies may concern five small corrections or one problem that still prevents use. No message may mean independent work or files put aside. Contact is evidence about communication. Its relationship to the customer's task determines whether the provider needs to act.

A completed service can leave a route open

Basecamp 5 documentation, updated on 24 October 2026, distinguishes archiving from deleting a project. Archiving removes it from the main view while preserving data and access through the project directory. It is an example of a tool feature: current work can close without pretending its history has disappeared.

Likewise, name what remains after completion. Does the customer know which version they accepted, where it was delivered and how to report a problem within the agreed scope? These help close a project. They do not create a new ongoing service or prove monthly customer activity.

Clarify later needs too. Correcting a discrepancy in accepted materials, adding a product line and producing new materials after a company change are different matters. The customer needs a route to raise them; the provider needs to recognise which part of the relationship each belongs to before calling every message a "return" or every silence a "departure".

First use of a service result may need a separate check if that stage belongs to the agreed work. It does not mean extending the project merely to stay in touch. Name the transition, such as a designated person independently using the materials. Once it is complete, the team can consciously mark the task done. For related guidance, see the article on first using a service result (in Polish).

Measure the relationships you actually compare

Separate one-off assignments from recurring services in the report. With recurring updates, investigate why a customer stops using them. With a one-off set, you mainly need to know about completion, use and remaining problems. The two models need not have the same order rhythm. Otherwise the studio may sell continued contact instead of necessary work.

The same company can use both models at once: a completed catalogue for one line and monthly updates for another. Marking the entire company active does not show whether the first catalogue fulfilled its task. Marking it lost when that project closed would ignore the ongoing service. Separate scopes before combining their results at customer level.

Similarly, define a new order. Sending a missing file, clarifying a name and commissioning a new set are different activities. If all increase a return counter, many corrections can look like desirable continuity. The event chosen should answer the question you need to resolve, rather than forcing work into the easiest system field.

Be especially careful about rewarding an account manager solely for the number of active projects. Leaving a case open longer may improve a report without helping the customer. Measures that encourage unwanted behaviour (in Polish) begin here with a concrete choice: should the employee help finish the task or preserve an active status?

Do not replace retention with one equally broad new number. The share of accepted sets does not explain every outcome either. Choose a small set of information suited to the service: agreed scope, confirmed acceptance, known state of first use and open difficulties, for example. Collect only what the company can reliably obtain and use.

A B2B relationship review can assess a particular relationship when there is a reason. It need not lead to a subscription. The right finding may be that the materials meet the present task, the customer knows how to contact the team after a change, and no more work is planned without a new need. This outcome deserves a record without a red alarm label. The guide to reviewing a B2B relationship (in Polish) explores this question.

An offer should allow a good ending

On the website, explain the completed result a customer can order and what follows handover. If updates are also offered, show the additional scope and when it may be needed. Do not present it as the mandatory continuation of every project when the customer can use the finished set independently. A helpful offer need not turn each solved problem into an endless service.

Aeneia Design can help turn these decisions into a service description and a contact route. The site should let a customer choose specific work and understand how it ends. That design task must rest on the company's actual service model. Rearranging buttons alone cannot explain why someone should return after completing a task.

A customer returning after a long gap needs their current needs understood. The old set may still be right for its original use while the company has a new product, audience or working method. This is a separate starting point. There is no need to pretend the old task continued or suggest that the earlier silence was a problem. See the article on a returning customer (in Polish) for a related perspective.

At the next report review, choose one "lost" customer with a one-off assignment. Reconstruct what they ordered, what state was confirmed and what remains unknown. Only then decide what to do. It may be clarifying unfinished work or closing a job well done. The metric should help distinguish those outcomes, not take offence at a customer who no longer needs help.

Aeneia / Aeneia editorial

Editorial article. Source links are included in the text.

About Aeneia

Brands we have worked with

  • L’Oréal
  • Garnier
  • Netflix
  • Carrefour
  • Warner Bros. Discovery
  • Biedronka
  • HBO Max
  • Wolt
  • Żabka
  • COSRX
  • Embassy of China in Poland
  • Mixa
Aeneia / Initium

From knowledge to a decision

Tell us what you want to put in order. We will agree on where to start.

Contact Aeneia

We partner with and use solutions from:

  • Google
  • OpenAI
  • Microsoft
  • Stripe
  • Anthropic
  • Linear