# B2B account-weighting and follow-up exercise

All people, accounts and records are fictional. No vendor has been tested.

## Reproduce the article's two calculations

Use fictional-account-responses.csv only. There are 12 distinct responding contacts:
Account A has 10 responses, 8 confirming resolution. Account B has 2 responses, none confirming resolution.

- Contact-weighted result:8/12 =66.7% when rounded to one decimal.
- Within-account rates:A=8/10=80%;B=0/2=0%.
- Equal-account result:(80%+0%)/2 =40%.

The first describes responding contacts; the second gives each responding account equal weight.
Neither describes all invited customers, proves an account's issue is resolved, or predicts retention.
No invitation roster or revenue weights are provided. Do not calculate response coverage, population
estimates or commercial impact from this file. The article's 20-account/60-contact coverage example
is a separate illustration and must not be combined with this 12-response dataset.

## Review what happened after feedback

Use separate-follow-up-scenarios.csv as a discussion worksheet. Its four scenarios are not additional
survey responses and must not be appended to the weighting dataset. They test disagreement,
missing confirmation, uncertain matching and a changed employer. They do not certify any product.

Before using real records, agree on the reporting unit, time window, missing-data rule, permitted
access, question wording and what qualifies as customer confirmation. Preserve source evidence and
separate staff status from customer testimony. Do not infer a person's identity from anonymous data.

## Proposed website use

Add one short download paragraph after the existing account-weighting explanation on
https://www.sopact.com/use-case/b2b-customer-feedback after review. Keep the current design and article.
Public asset hosting, a current CMS insertion check and exact publication approval remain pending.