The essentials
Match the kind of claim to its evidence
Start with the uncertainty your reader needs to resolve. A service manager may want to know whether you understand handovers between dispatch and engineering. A founder may want to see the actual deliverable. Choose one uncertainty, then find material that answers it. A familiar logo, a generous compliment, and a measured case result communicate different things.
Build the sentence from the evidence you have. A completed worksheet can demonstrate your method. An approved customer quotation can describe that person’s experience. A case record can describe a measured change within its documented scope. Keep those roles distinct when editing: a positive quotation does not supply missing measurements, and a polished example does not establish a customer outcome.
This guide provides an editorial method for assembling and checking that material. The FTC’s substantiation policy is a primary reference for having a reasonable basis before making an advertising claim. Its US policy context does not establish a universal rule for every email. The practical workflow here is to inspect the support before writing the result sentence.
- Write the reader’s uncertainty in one sentence: “Can this team make an engineer handover usable?”
- Name the available support: an approved case record, a specific quotation, or a demonstrable work sample.
- Draft the narrowest useful statement the material can substantiate. Add ambition to the offer separately from the evidence.
References [2]
A closer look
Open one fictional case packet
The following NorthQuay Service packet is invented to teach the method. NorthQuay is an equipment service team, and none of these figures, approvals, or people represents a real customer. Do not reuse its result as evidence for your business. Keep the packet together while editing so each shorter version can be checked against the same record.
The operations lead has compared handover records before and after a checklist and staff training were introduced on 15 September 2026. A handover is complete when four fields are present at dispatch: customer or site, asset ID, fault summary, and owner. The record measures completeness at that moment. It contains no measure of repair speed, revenue, or customer satisfaction.
| Packet field | Recorded detail |
|---|---|
| Baseline period | 1–14 September 2026: 43 total handovers; 3 cancelled records excluded; 40 eligible; 26 complete. |
| Comparison period | 15–28 September 2026: 45 total handovers; 5 cancelled records excluded; 40 eligible; 34 complete. |
| Method | Count eligible handover records, not people. Apply the same four-field definition at dispatch and the same cancellation exclusion to both periods. |
| Intervention and limit | Checklist and training introduced together on 15 September. An observational comparison; other differences between periods were not controlled. |
| Record and cut-off | Operations lead owns the export and definitions. Export reviewed through 2 October 2026. |
| Permission and review | Fictional approval dated 6 October permits anonymous counts and method in one-to-one email and a case PDF. No logo or named-quotation permission. Review on 6 November 2026. |
A closer look
Keep dates and denominators attached
Calculate the rate from the eligible records in each period. Here the exclusions are applied before division: 43 minus 3 gives 40 baseline records; 45 minus 5 gives 40 comparison records. Show the starting totals and the reason for exclusions in the supporting case so a reviewer can reconstruct both denominators.
Use percentage points for the difference between the two rates. Relative change answers another question: how large the increase is compared with the starting value. The arithmetic can be correct while the wording is wrong. Neither calculation measures speed, so “31% faster” would invent an outcome the packet never recorded.
Before accepting a real comparison, check whether the definition and collection method stayed consistent. If the second period records missing fields more carefully, or the eligible work is materially different, explain that difference. If comparable data is unavailable, describe the observed records without presenting a before-and-after improvement. Choose the form of the claim after reviewing the data.
| Measure | Calculation | Usable interpretation |
|---|---|---|
| Baseline completeness | 26 ÷ 40 × 100 = 65% | 26 of 40 eligible handovers had all four fields. |
| Comparison completeness | 34 ÷ 40 × 100 = 85% | 34 of 40 eligible handovers had all four fields. |
| Absolute rate difference | 85% − 65% = 20 percentage points | The observed completeness rate was 20 percentage points higher. |
| Count difference | 34 − 26 = 8 records | Eight more complete handovers, with equal eligible totals. |
| Relative rate increase | (85 − 65) ÷ 65 × 100 ≈ 30.8% | A relative increase in recorded completeness; no measurement of speed. |
Worked example
Turn the packet into a bounded statement
Build a full source statement first. Include the type of team, measured outcome, dates, counts, and what changed. Describe the intervention alongside the observation. The NorthQuay record cannot isolate the checklist’s contribution because training began at the same time and other period differences were not controlled.
Next, make an email-length version by removing background that does not change the meaning. Preserve the essential boundaries: one team, eligible handovers, the comparison periods, and the observed result. Put the detailed definition and exclusions in the supporting case. If removing a qualification changes what a reasonable reader would understand, keep it beside the claim.
- Replace “teams” with “one team” when the evidence covers one case.
- Name the actual measure: completeness at dispatch, rather than efficiency or performance.
- Keep the statement historical. A previous result does not specify what this recipient will achieve.
- Check the final sentence against the approved version after every sales edit. A changed verb can change the claim.
FIRST DRAFT
Our proven system makes service teams 31% faster. We transformed NorthQuay’s operations in two weeks.
A MORE FOCUSED VERSION
In one equipment service team’s September 2026 comparison, complete handovers rose from 26 of 40 eligible records to 34 of 40 after a checklist and training were introduced. The periods were 1–14 and 15–28 September. This records a change in completeness; it does not isolate the cause.
The revision replaces an unmeasured speed claim with the observed metric, removes an unauthorised name, and identifies the comparison and attribution limit. In a real campaign, every part would need real evidence and appropriate permission.
Worked example
Place a compact case example in a full email
Use the case to explain why the offered next step is relevant. Give the reader enough context to recognise the connection, then offer the supporting material you can actually supply. Avoid adding several unrelated results. One case with a defined measure gives you a coherent reason to send one useful follow-up.
The two messages below belong to the fictional NorthQuay exercise. The recipient and sender are invented. They model how a real, approved case might be used; they are not available customer proof. Notice that the follow-up supplies the definition and exclusions promised by the first email.
Illustrative first email
An invented provider writes to a service operations lead about handover completeness.
SubjectHandover fields at dispatch
Hi Priya, Your service page describes engineers covering several customer sites. I help service teams define the information that needs to travel with each dispatched job. In one team’s September 2026 comparison, complete handovers rose from 26/40 eligible records to 34/40 after a checklist and training were introduced. That was a completeness measure, not a controlled test of the checklist. Would the one-page case, including the field definition and exclusions, be useful? Alex
Relevant context
The public observation connects to the work being discussed without claiming knowledge of the recipient’s internal problems.
Specific support
One team, one month, and visible counts keep the example bounded. The supporting page supplies the exact periods.
Fulfillable action
The requested reply leads to a prepared case page with the promised method.
Illustrative fulfilment reply
The recipient has asked to see the case. The case page described here is part of the exercise.
SubjectRe: Handover fields at dispatch
Hi Priya, Here is the one-page case. “Complete” means customer/site, asset ID, fault summary, and owner were all present at dispatch. For 1–14 September, 26 of 40 eligible handovers were complete; for 15–28 September, 34 of 40 were complete. The periods contained 43 and 45 total records respectively, with 3 and 5 cancelled jobs excluded. The checklist and training began together on 15 September. Other differences between periods were not controlled, so the comparison does not isolate their effect. The export was reviewed through 2 October. The final section shows the four-field checklist used in the comparison. Alex
Complete the handover
Provide the promised document in the actual sending workflow. Check its access, version, and permissions before using this structure.
Consistent figures
The follow-up preserves the same counts and dates and explains how the eligible totals were formed.
Worked example
Preserve permission and quotation meaning
Treat permission as specific to the material and its use. For the fictional packet, the approval covers anonymous counts and the method in one-to-one email and a case PDF. It does not authorise a company logo, a named testimonial, or a public website case. Prepare the anonymous version within that scope; seek a separate approval if you want a different use.
For a quotation, retain the original wording, who supplied it, their role at the time, the date, and the approval record. Ask the owner to approve the exact edited quotation and its attribution. If you paraphrase, present it as a paraphrase. Do not turn a comment about a useful workshop into a numerical result or imply that a person endorsed services they did not discuss.
ASA/CAP’s testimonials guidance provides a useful primary reference for checking that testimonials are genuine, relevant, accurately presented, and used with appropriate permission. This is UK advertising guidance. Here it informs an editorial review habit; it is not a determination of the rules applying to a particular campaign.
FIRST DRAFT
“The new worksheet gave dispatch and engineering the same fields to check.” → “This transformed our efficiency.”
A MORE FOCUSED VERSION
Keep the original sentence if that exact wording and attribution are approved. If space is tight, ask the speaker to approve a shorter version that preserves the point about shared fields.
A genuine process comment can be useful on its own. Adding an efficiency claim changes its meaning and requires different support.
References [1]
Put it into practice
Use the evidence you have and maintain it
If you have no approved customer result, show a process sample. Build a clearly labelled fictional handover, explain why each field matters, and offer the blank worksheet. Your claim can be “Here is how I structure a dispatch handover.” That lets the reader inspect your thinking without presenting a made-up outcome as customer experience.
For a real historical case, set a review date and an owner. Recheck whether the approved wording, sharing scope, service description, and supporting document still fit the campaign. A historical result can remain useful when its dates stay visible. It becomes misleading if an old snapshot is silently presented as the current customer experience.
At the fictional packet’s 6 November review, the editor would check the permission and source file before reuse. If a count changes, update the calculation, email sentence, case PDF, and saved reply together. Keep the previous version and a short change note so colleagues know which statement is ready to use.
Put it into practice
Copy the proof worksheet
Complete this for one statement before adding it to a sequence. Read the email sentence aloud, then trace each factual element to the packet. Ask a colleague to check the calculation and explain the claim back to you. Resolve any difference in meaning before using it.
Case statement and supporting evidence brief
Copy this worksheet into your own document, then fill it in with your details.
READER AND PURPOSE Reader’s specific uncertainty: Why this case is relevant: EVIDENCE PACKET Case owner and source location: Customer or team description approved for use: Metric and exact definition: Baseline dates / total / exclusions / eligible / result: Comparison dates / total / exclusions / eligible / result: Calculation, units, and rounding: Intervention and other relevant changes: Collection method and limits: Export or evidence cut-off date: PERMISSION Approver and approval date: Exact approved wording or quotation: Approved attribution, name, and logo use: Approved channels and supporting document: Permission record location: READY TO SEND Full source statement: Email-length statement: Qualification that must stay beside it: Supporting asset and fulfilment reply: Owner / review date / version: Reviewer’s calculation and meaning check:
Copy the brief into your own workspace. This page does not save a brief or submit campaign information.
MAKE THE EVIDENCE CHECKABLE
Choose a relevant case, preserve its dates and denominator, and keep the approved statement connected to the material that supports it.
A LITTLE MORE CLARITY
Your questions, answered.
What social proof can I use if I do not have customers yet?+
Use an inspectable work sample and describe it accurately. A fictional example, annotated worksheet, or explanation of your method can show what you would deliver. Label it as a sample and make claims about the work itself. Reserve customer-result statements for real cases with support and permission.
Can I use an anonymous customer case?+
Yes, when your evidence and sharing permission support that use. Retain enough approved context to make the case meaningful, such as the team type, measured work, and dates. Check whether the remaining details could still identify the customer. Keep the source and approval internally so anonymity does not remove traceability.
Should I write percent or percentage points?+
Use percentage points for the difference between two percentages. Moving from 65% to 85% is an increase of 20 percentage points. Relative to the original 65%, it is about a 30.8% increase. Name the metric in either version and show the underlying counts when they help the reader understand the comparison.
Can an older case still be useful?+
A dated historical case may still illustrate relevant work. Review its permission, method, and fit with your current offer before reuse. Keep the measurement dates visible, and avoid implying the result is current. If the supporting document or wording has changed, update every place where the claim appears.