Win Work You Can Deliver

Chapter 9 of 10 · 5 min read

Earn the next conversation

Review delivery honestly, handle problems according to their cause and obligations, and make follow-up useful rather than automatic.

The end of a project is a chance to learn whether the work helped, where the process failed and whether a future conversation would be useful. It is not automatic permission to start selling again. A customer may be satisfied and have no further need; another may need help but be reluctant to raise a problem.

Fieldnote has delivered Harbour Services’ approved documents. The files exist, but that alone does not establish that staff use them or that misunderstandings have decreased. A proportionate review should examine those questions without claiming more evidence than is available.

Review use as well as delivery

Agree a suitable review point based on when the customer can use the output. Asking for an outcome immediately after handover may be premature. Waiting indefinitely loses useful detail. For this example, the internal training session creates a sensible opportunity to ask what happened in practice.

Fieldnote asks which document staff used, where they still needed clarification and whether anything important was missing. It does not begin with “How much time did we save you?” because no dependable baseline has been established. If the customer reports an improvement, record the observation and its context rather than converting it into a general performance claim.

Compare what you can actually observe with the original purpose. The service rules may now be consistent, while the customer-facing summary is rarely used because staff cannot find it. That is a different problem from poor wording. A useful review distinguishes the quality of the output, its implementation and external influences.

Figure 1Closing the learning loop
  1. What was promisedthe accepted offer
  2. What was deliveredversions, changes, handover
  3. How it was usedobserved after a suitable interval
  4. What we learnedeffort, input quality, usefulness
  5. Next decisionselection, pricing, delivery, follow-up
Delivery evidence feeds back into which work to choose, how to price it and how to deliver it.

Treat problems as decisions with consequences

When a customer reports a problem, first establish what happened, what is affected and whether an immediate action is needed to limit harm. Acknowledge the report without making unsupported conclusions about cause or liability. Give a clear response time and keep the customer informed if investigation takes longer.

In the fictional case, Harbour Services finds an old package name in a final document. Fieldnote checks the accepted inputs and discovers it introduced the error during a late edit. The team acknowledges the mistake, supplies a checked correction, identifies the affected version and confirms where it was distributed. It also improves its final consistency check.

Contrast that with a new package introduced by the customer after handover. The document may now need updating even though the original work met the approved requirement. Explain the difference with reference to the agreed scope and chronology. Avoid treating every complaint as extra work or every new requirement as a supplier failure.

Figure 2Two reported problems, two different responses
Old package name in a final documentNew package added after handover
SourceWithin the agreed scopeCustomer change after completion
ResponseCorrect it promptlyOffer an update as new work
RecordFactual note of the issue and fixChange or new offer
Establish what happened and what is affected before deciding whether the issue is an obligation or a new request.

Separate obligations from discretionary goodwill

Australian consumer guarantees can apply to qualifying purchases, including some business purchases. The ACCC’s consumer rights and guarantees guidance explains that these rights arise automatically where applicable. A remedy owed under law is not merely a goodwill favour. Check the circumstances and obtain advice when necessary rather than relying on a blanket “business customers are excluded” assumption.

After addressing applicable obligations and the agreed scope, the business may choose additional assistance. Record what it is offering and why. A modest help session might be worthwhile to clarify use; an open-ended promise to fix any future change is a different commitment. Generosity is easier to sustain when its boundaries are understood.

For an unresolved disagreement, maintain a factual record of the issue, relevant documents, proposed resolution and next step. Avoid using the prospect of a negative review as the sole reason to act. The decision should be defensible for a customer who never posts publicly.

Ask for feedback without scripting the answer

A useful feedback request permits criticism. Ask what worked, what required too much effort and what the customer would change about the process. An invitation that only welcomes praise will produce a flattering but incomplete picture.

Do not turn private feedback into a public testimonial without permission for that use. Confirm the wording and context, and do not add outcomes the customer did not describe. If you request public reviews, use a consistent process that does not pressure customers to give a favourable rating or conceal dissatisfied experiences.

Fieldnote can learn from one customer’s comments without treating them as representative of every buyer. If three customers mention confusing review instructions, that pattern justifies investigating the instructions. It still does not establish a precise percentage of the market that experiences the problem.

Measure repeat work with a meaningful denominator

Repeat purchasing depends on the service and observation period. A one-off document package may not need to be bought again soon. Counting every completed customer as a failed repeat sale after thirty days would misrepresent the business.

Suppose ten fictional customers completed work in a period. Six have subsequently reached a defined annual update window, and four of those six commission an update. The observed repeat-purchase proportion among that eligible group is 4 ÷ 6, or about 66.7%. The other four customers are outside this particular observation window. State the rule and the small count alongside the percentage.

Figure 3Repeat purchases with an honest denominator
Completed customers10
Reached the annual update window6 eligible
Commissioned an update4 66.7% of eligible
4 of 6 eligible customers, about 66.7%. The other four are outside this observation window, not failures.

Do not quietly redefine eligibility to exclude customers who chose not to return. Establish the rule before examining outcomes, and separately record cancellations, uncontactable customers and known changes in need. A measure that improves only because inconvenient cases disappear is not useful learning.

Make follow-up serve a current need

A follow-up can be tied to an agreed review date, a known service change or a requested update. State that reason. Ask whether the conversation remains useful, and respect the customer’s response. Permission to complete a project does not establish blanket permission for unrelated marketing.

For Harbour Services, the next useful conversation may be before its planned annual package review. Record that timing if the customer wants it. If no continuing need exists, close the engagement clearly and make it easy to return later. A respectful ending is a valid commercial outcome.

Exercise: close the learning loop

Review one completed engagement using four questions: What was promised? What was delivered? What happened when the customer used it? What should change in the next offer or delivery process? Assign one improvement to a person and date. Keep observations separate from explanations you have not tested.

The customer relationship now feeds evidence back into selection, pricing and delivery. The final chapter assembles these practices into a manageable thirty-day implementation.