Run a Vendor Demo With Your Own Shipment Data
A presentation-style evaluation guide for freight buying teams that want vendors to run the same shipment scenarios and produce comparable evidence.
· AI-assisted editorial content

A software demo becomes more useful when the team supplies a shipment case that exposes actual decisions. An export clerk should be able to book against the original instructions, accept a change only after the right approval, and find what was sent later. A finance colleague should see whether that change alters the basis for invoicing. The demo is a test of those decisions across the workflow.
This presentation uses a fictional sea-export case. Make your test pack from approved synthetic or appropriately de-identified material and arrange access with the vendors before sharing. Give each supplier the same brief, dataset, time, and scoring rules so the buying team can compare evidence from one demonstration to the next.
Slide 1 · Choose the decisions you will watch
List the few operational decisions that could change your purchase. For example, a team might need to preserve which version of a shipping instruction was approved, check that the dispatch desk sees a roll notification, and stop an invoice when the debit terms conflict with the quote. Write one sentence per decision and identify the job function that owns it today.
For every decision, define something a moderator can observe. A practical criterion is whether the operations user finds the latest approved instruction from the job record and whether a read-only viewer sees the same active revision. An interesting dashboard demonstration can still receive a low score if those expected behaviors are missing. Keep separate entries for required controls and capabilities you would be willing to defer.
Slide 2 · Build a small, coherent shipment pack
Create a few source documents that tell a consistent story. Include a quote, a booking confirmation, the original shipping instruction, a draft ocean bill, a revision, and a simplified cost record. Keep their dates, pieces, parties, and amounts internally consistent until the moment when the script introduces a controlled change. Use fictitious contact details, company names, and identifiers when creating the files.
Retain only the complexity that affects the decision. One extra consignee with a similar name can show whether the data entry screen distinguishes parties. Ten almost identical contacts may instead waste demo time. Number the inputs and write a separate, short expected-output sheet that your evaluators can use. Vendors can receive the base pack in advance. Hold a synthetic change request for the live session and tell everyone that a change will be introduced.
Slide 3 · Set a common demo environment
Ask the presenter to identify the environment, user roles, and connection status. A configured demonstration tenant can be appropriate when the same assumptions are available to each supplier. Mark values already imported and record which partner integrations are simulated. If the demo relies on an approver who works in a separate tenant, include that condition in the test notes.
Assign someone from the freight team to moderate. Give them a written sequence and a fixed time allowance for each decision. An observer should track evidence and another should run the stopwatch if response time actually matters. Confirm vendor permissions for any saved screenshots or output. Identify precisely which copy of the test pack was supplied, and allow each vendor the same setup and troubleshooting time.
Slide 4 · Introduce one controlled revision
Start by asking the demonstrator to process the original shipping instructions, verify the resulting draft, and show the two user roles that can view or approve it. Then hand over a fictional instruction that changes one named consignee detail after approval, with an approval already supplied through the test script. Have the team mark the expected impact on the draft and the shipment timeline before the vendor acts.
Ask the operations user where the active version can be seen, how the older version is retained, and whether an approval is needed before the revision goes externally. Ask finance whether the changed field affects the invoice and what evidence confirms its charge basis. When a demo succeeds only because an administrator edits a record behind the scenes, score that dependency explicitly and distinguish it from a workflow available to the desk.
Slide 5 · Score what actually happened
Keep a simple scale for the operational behaviors. A zero can mean no evidence, one can mean a workaround or an unproven dependency, and two can mean the expected result was demonstrated under the agreed roles. Record a short narrative describing the result regardless of the score. A buyer considering a manual workaround needs to know who will perform it and how often.
Write the source of any uncertainty. An export that was promised but absent may be pending evidence. A feature visible only with a paid add-on may require a pricing clarification. A controlled test run that was interrupted by an environment failure may be repeatable under identical conditions. Ask the moderators to compare scores after they have each entered their own evidence, so a persuasive presenter does not set the first impression for the whole room.
Slide 6 · Decide the next proof you need
List the mandatory behaviors with no convincing evidence and request a narrowly scoped follow-up. Give vendors the same chance to provide a missing output under the agreed test conditions. Connect commercial questions to specific dependencies that were demonstrated. If an attachment export requires a support ticket or a charge rules feature requires a different plan, ask for that dependency in writing.
Retain the outcome as a structured record. Capture the input version, test environment, observed behavior, score, and unresolved question. Reuse this record when negotiating implementation and when designing user-acceptance tests. Tasks that were only simulated during buying can then become an explicit acceptance step rather than a vague expectation inherited from a meeting.
Test a revised consignee instruction in 25 minutes
The fictional Bayline team tests shipment Q-027. The approved draft shows a consignee billing office, and a subsequent instruction changes its receiving office. The booking and sell rates are unchanged. The team wants the old and revised instructions preserved in the job, the new draft routed for approval, and the quote amount retained on the billing record.
Supplier A shows both revisions and opens the approved quote from a billing account, but the facilitator has to trigger the review workflow through an administrator. Supplier B demonstrates the approval through the configured desk role but says the audit history can be exported later. The team records a separate unanswered question for each supplier and requests the same evidence window for follow-up.
| Decision | Expected evidence | Record if missing |
|---|---|---|
| Find the active instruction | Revision number and approval user visible from job | How an operator currently identifies the active copy |
| Approve as desk user | Approval works under supplied desk role | Setting or role needed to resolve the gap |
| Check the quote basis | Sell rate and effective quote stay visible to finance | Whether vendor or finance staff own the correction |
| Retrieve history | Earlier instruction and event record can be reviewed | A sample export and any plan dependency |
Desk checklist
- Write criteria and expected evidence before the vendors arrive.
- Use the same synthetic pack, roles, and timebox for each demonstration.
- Choose a revision that tests two working teams rather than one screen.
- Record what was demonstrated and what was simulated.
- Keep evidence and workaround explanations beside every score.
- Carry unresolved mandatory controls into purchasing and acceptance tests.