Improve proposal production

For Haruka Works / A four-week pilot proposal

This is an illustrative proposal for a fictional company. Customer information, business results, metrics, costs and expected benefits are invented. They are not real company results or evidence of WworPoint's effectiveness.

Executive proposal

We consider a fictional situation in which growing inquiries put pressure on the time available to create proposals. For four weeks, test a workflow that keeps content, evidence and design in files, with a consistent path from review to export.

The aim is not only to work faster, but to make evidence and changes easier to check.

The assumed challenge

The baseline assumes 12 working hours per proposal. Transcribing information, adjusting layout and comparing finished files take time. Improvement may be possible, but the team first needs a shared measurement method.

Scope and participants

Use one proposal theme and its reviewers. Manage Markdown, local CSV/JSON, CSS and visual definitions. Moving customer information outside the organization or connecting business systems is outside this example's scope.

Requested decision

Approve a four-week pilot and set aside time for weekly reflection. Consider wider adoption only after reviewing the results.

01 / Share the challenge with data

Inquiry and proposal trends

Illustrative monthly inquiries and proposals

From April to September, fictional inquiries rise from 120 to 240 and proposals from 52 to 78. These figures alone do not establish a production bottleneck. Check inquiry quality, staffing and case complexity too.

Working time by task

Illustrative task hours and untested pilot targets

A target of 7 hours against a 12-hour baseline implies 5 hours, or about 42%, less time. This is an untested pilot hypothesis, not a promised reduction.

Measurement assumptions

Compare proposals of similar difficulty. Record working time from the first draft to approval. Separate waiting and meetings, and retain revision details so that differences in quality are not overlooked.

02 / Connect evidence and review

Proposal production process

The author drafts the content; reviewers check facts, evidence and the proposed decision. If AI tools support the draft, people still verify and revise it before submission.

Structure of the deliverables

Manage text in Markdown, data in local CSV/JSON and design in CSS. Generate visuals from Vega-Lite and Mermaid, then validate and export PDF and static HTML with WworPoint. No built-in AI service or editing GUI is assumed.

Review responsibilities

Check	Responsible people	Acceptance criterion
Figures and sources	Author + reviewer	Matches source data
Proposed decision	Business owner	Assumptions and exceptions are clear
Layout	Author	Validation and PDF reviewed

Check the separate network policies of any AI tools or editors used.

03 / Pilot, measure and decide

Four-week implementation plan

Illustrative pilot plan
Week	Milestone	Owner	Check
1	Baseline agreed	Sales ops	Measurement method
2	Test one proposal	Sales rep	Content and evidence
3	Fix review steps	Sales lead	Changes and reproducibility
4	Review and decide	Pilot team	Quality maintained

Evaluation measures and decisions

Measure	How to record it	Decision after the pilot
Working time	Hours by task	Less time at equivalent quality?
Rework	Revision count and reasons	Easier evidence checks?
Content quality	Common review checklist	No increase in errors or gaps?
Sustainability	Team reflection	Acceptable learning effort?

Risks and responses

Match figures to source data. People verify AI-generated text. Adapt the design to the reader and purpose. Passing layout validation does not establish that the content is correct.

The next decision

In week four, share the records and decide whether to continue, improve or stop. Expand gradually if continuing. Costs and benefits are not established, so this example does not calculate financial ROI.

All figures are fictional. In real work, measure assumptions and outcomes before deciding.

04 / Compare the pilot assumptions

Landscape appendix / Fictional working-time comparison

A wider page keeps task hours, evidence and review criteria side by side. The 12-hour baseline and 7-hour target are illustrative assumptions, not measured results or a promised saving.

Illustrative assumptions and review checkpoints (hours per proposal)
Task	Baseline h	Pilot target h	Difference h	Evidence to retain	Reviewer	Acceptance check
Research	3	2	1	Source list and case brief	Author + reviewer	Check sources and scope
Drafting	4	2	2	Draft and revision notes	Author + reviewer	Check facts and assumptions
Review	3	2	1	Completed review checklist	Business owner	Decision and caveats clear
Layout	2	1	1	Validated PDF and source	Author	Readable complete export
Total	12	7	5	Comparable-case work logs	Pilot team	Equivalent quality first

Read the comparison together

Record actual task hours during the pilot and retain the evidence shown above. Compare cases of similar difficulty. Decide after checking quality and reviewer feedback; do not treat the 5-hour difference as established performance.

All figures are fictional. This PDF contains four A4 portrait pages followed by one A4 landscape page.