PortproofThe item rose. The total fell.
A comparable observation includes delivery and the conditions behind the offer.
| Price component | Run A | Run B |
|---|---|---|
| Item | €24 | €25 |
| Delivery | €4 | €0 |
| Delivered total | €28 | €25 |
Change in delivered total
The item price is €1 higher.
The delivered price is €3 lower.
- Product and quantity
- Same variant, one 500 ml item
- Seller and availability
- Same seller, available in both runs
- Currency
- EUR in both runs
- Destination
- Same saved delivery destination
- Tax and promotions
- Same treatment and conditions
- Storefront context
- Same locale and account state
Missing delivery is unknown, not free. Run B explicitly states €0 delivery. Preserve missing shipping, tax or stock fields for review.
No retailer requests or live prices are used.
A response is not a valid price.
In this separate fictional observation, the acceptance rule requires EUR.
Required currency
EUR
Returned currency
USD
Keep the attempt and its reason. Do not save it as a new comparable price or silently convert the currency.
The traffic still counts.
Include transferred bytes from rejected attempts and retries once in the sample total. Do not add a second retry multiplier.
Keep the comparison context.
Record the source URL, capture time, required fields and rejection reason. Keep credentials and personal data out of the record.
Your collector validates the result. A proxy connection does not choose the storefront, currency, tax treatment or delivery postcode.
This illustration performs no collection.
Start with the whole sample.
Scale all sample bytes by accepted observations and the planned schedule.
These 120 accepted observations are a separate invented sample. They did not come from the offer examples.
- All sample bytes
- 10,800,000
- Accepted observations
- 120
- Products
- 1,000
- Markets per product
- 2
- Checks per product daily
- 4
- Days
- 30
Fictional forecast with 20% headroom
Decimal GB: 1 GB = 1,000,000,000 bytes.
- Bytes per accepted observation
- 90,000
- Planned observations
- 240,000
- Before headroom
- 21.6 GB
- Chosen allowance
- 4.32 GB
Forecast, not a bill. Headroom is an allowance for variation, not a guarantee or measured confidence interval.
All sample bytes include captured uploads, responses, redirects, unsuccessful attempts and retries, counted once within the counter’s scope. Browser transferred and uncompressed sizes differ; neither is automatically the proxy billing meter.
No accepted observations means no per-result estimate.
Decide before increasing the run.
For your own permitted trial, write acceptance conditions first. One passing check cannot cancel another required failure.
Go
Required results are correct. Task times and settled, isolated usage fit your preset budgets.
Plan the next amount from your measurement. Recheck after changing the workload.
Hold
Usage has not settled, other workers share the window, or representative cases remain untested.
Complete the measurement. Unsettled usage is unknown, not zero.
Stop
A required result is wrong or missing, or a task exceeds its deadline.
Investigate and retest required tasks before increasing the order.
No account needed for the planner.
This link opens the live Portproof website.
Use your own representative, permitted sample. Compare your counter with settled account usage. A sample cannot guarantee future country availability or performance. Checklist selections stay in this page and reset when it is reopened.