Units
1. Introduction to Artificial Intelligence in Project Management: Roles, Boundaries, Authentication and Privacy 2. Work Breakdown Structure (WBS) and Scope Planning 3. Time and Resource Estimation: Time Estimation and Verification 4. Project Schedule: Milestones, Dependencies, and Critical Path 5. Risk Management and Risk Registration 6. Status Report and Progress Monitoring: EVM, Deviation and Executive Summary 7. Meeting Summary and Action Tracking 8. Stakeholder Communication and Expectation Management 9. Change, Problem and Quality Management 10. PMO Documentation and Information Management 11. Data Privacy, Ethics and End-to-End Governance
Unit 6 / 11

Status Report and Progress Monitoring: EVM, Deviation and Executive Summary

Gains:

  • Ability to understand status report components and earned value (EVM: PV, EV, AC, SPI, CPI) indicators and produce a management summary draft with the support of artificial intelligence
  • Ability to use AI for deviation explanation, trend interpretation and red-yellow-green (RAG) status sketching
  • Understanding that it is up to the project manager to verify the numbers and status colors in the report with real project data and to recognize the risk of optimistic reporting (watermelon).

The most frequently repeated task after a project begins is monitoring and reporting status. A status report is a regular document that summarizes the health of the project at a particular point in time to stakeholders: where we are, how we are doing according to plan, what risks exist, what decisions are needed. A good status report is honest, concise, and decision-oriented; A bad condition report is either a pile of numbers or a veneer that hides the truth. AI is very powerful at speeding up status reports: it takes raw progress data, turns it into a neat executive summary in minutes, writes draft sentences to explain deviations. But this very power carries a trap: AI can reinforce optimistic input with polished language and make an actually problematic project look “good.” It is up to the project manager to verify status colors and numbers with reality.

Earned value management (EVM) basics

Measuring progress with subjective expressions such as "what percent done" is misleading. Earned value management (EVM) is a method that objectively measures progress in terms of both time and cost. Let's define three core values:

  • PV (Planned Value): The budget value of the work that should have been done according to the plan by this moment.
  • EV (Earned Value): The budget value of work actually completed (percent complete × budget).
  • AC (Actual Cost): The amount actually spent to do this job.

Two powerful indices emerge from these three:

  • SPI (Schedule Performance Index) = EV / PV. Under 1: behind plan. Above 1: ahead of plan.
  • CPI (Cost Performance Index) = EV / AC. Below 1: budget overrun. Above 1: under budget.

Example: If PV = $100,000, EV = $80,000, AC = $95,000, SPI = 0.80 (behind schedule, 20% of work missing) and CPI = 0.84 ($0.84 in value for every $1, over budget). These two numbers replace feelings like “we're doing well” with concrete reality.

The situation is usually summarized by the colors RAG (Red-Amber-Green). Green: on track. Yellow: attention required. Red: intervention is necessary. What's dangerous is the watermelon effect: green on the outside, red on the inside — so the report says "green" but the project is actually problematic. This arises from the tendency to delay bad news, and the polished language of AI can unknowingly foster this.

indicator

formula

Meaning

health sign

SPI

EV/PV

Time performance

<1 behind, >1 ahead

CPI

EV/AC

Cost performance

<1 budget overrun

SV (deviation)

EV − PV

Time deviation (TL)

Negative = behind

CV (deviation)

EV − AC

Cost variance (TL)

Negative = overshoot

Step by step: Status report with AI

  1. Collect real data. Pull PV, EV, AC values ​​and mission progress from your own system. AI can't make these things up; you give.
  2. Calculate/verify indices. Ask the AI ​​to calculate SPI, CPI, SV, CV, then manually check at least one of them.
  3. Ask for an explanation of the deviation. "Why is SPI 0.8?" AI produces a draft explanation to the question; you verify the real cause (permission, dependency latency, scope escalation).
  4. Determine RAG status. You assign colors according to the data; Limit AI's optimistic language.
  5. Decision-oriented summary. Ask the AI ​​to clearly write the "what decision/request is required from the sponsor" section.
  6. Integrity audit. Before sending the report, "is this report really green or watermelon?" ask yourself.
Caution: Telling the AI ​​to “write the report in a positive tone” increases the risk of hiding bad news. The tone should be professional but not soften the truth. The sponsor prefers to hear bad news early rather than learn it late.

three mini cases

Case 1 — Fast and honest report. A PM would give the EVM data of 6 work packages to AI every Friday and receive a draft management summary. AI calculated SPI/CPI, attributed deviations to RAG. PM verified each number with his clipboard and corrected a statement. Report time decreased from 90 minutes to 20 minutes; quality increased.

Case 2 — Watermelon caught. One coordinator compared the AI's "overall green, minor delays" sketch with CPI 0.82 and SPI 0.78. The numbers were clearly yellow-red. The coordinator corrected the status to "yellow" and clearly wrote the reason for the delay. The sponsor intervened early; Two weeks later the report actually turned green. Lesson: color should match the data.

Case 3 — Fake percentage. “Write the completion rate in the report,” a team member told the AI, without giving the raw data. The AI ​​made up a number saying "73% complete"; the actual was 58%. The report gave false confidence to the sponsor. Lesson: EVM numbers should come from real data, not from the AI's memory.

Weak prompt / Strong prompt

Weak prompt:

Write this week's status report of the project in a positive language.

The request for “positive language” invites the watermelon effect; Also, since no data is given, the numbers are made up.

Powerful prompt:

Your role: assistant reporting assistant to a project manager.Context: Below is actual EVM data and package progresses (amounts as masked rate).Task: Write a draft weekly status report.Rules:- Calculate SPI, CPI, SV, CV values ​​from the figures I give and show the formula.- Recommend RAG color based on SPI/CPI thresholds (e.g. <0.9 yellow, <0.8 red), softening the tone.- ASK probable cause for each deviation, fitting; I will confirm the reason. - Write clearly the "Decisions required from the sponsor" section. - Do not make up any completion percentages. just use the data I give.Data: PV=..., EV=..., AC=..., packets: ...

This prompt is powerful: it includes real data, formula transparency, RAG threshold, tone warning, and fabrication ban.

Additional templates:

# Watermelon inspectorThe status text below says "green". Is this color consistent with the SPI/CPI and risk data I provided? If inconsistent, what color should it actually be and why? Say it clearly.

# Deviation explanatorySPI dropped to this value. List possible causes category by category (source, dependency, scope, quality) and write down what data I can verify for each. Why fake it?

# Trend summarizerExport SPI/CPI values ​​for the last 4 weeks; Interpret the trend (improving/worsening/stable) and suggest a range for estimated completion/budget (EAC) if the current pace continues.

Common mistakes

  • Asking for a “positive tone”: Softening the truth creates a watermelon effect and breaks down trust.
  • Making up numbers: EVM and completion rate should come from real data.
  • Color-data discrepancy: Saying "green" when SPI is 0.7 refutes the report.
  • Skipping the decision part: The report should say “what is needed,” not just “situation.”
  • Not validating indices: AI may make the calculation wrong; At least one of them must be manually controlled.
  • Too long report: Sponsor does not read 5 pages; One page of honest summary is worth more.
Tip: Set up each status report to answer three questions: Where are we (SPI/CPI)? What's the biggest risk? What is required from you? If these three are clear, the report has done its job.

In summary

The status report communicates the health of the project in an honest and decision-oriented manner. EVM (PV, EV, AC and derivative indices SPI, CPI) measures progress objectively, freeing it from subjective feelings. AI speeds up the report, explains deviations, and organizes the summary; but the optimist can polish the input, create a watermelon effect, and make up numbers if no data is given. It is up to the project manager to provide the actual data, verify the indexes, keep the colors consistent with the data, and deliver the bad news in a timely manner.

Application task

Give the actual PV, EV and AC data of your project (as a ratio, masking the amounts) to the AI and produce a draft status report. Verify SPI and CPI manually; You assign the RAG color according to numbers. Test color-data consistency with the "watermelon inspector" template and clarify the "decisions required from sponsor" section.

checklist

  • [ ] I took the EVM numbers from real data, I did not make them up.
  • [ ] I manually verified at least one of SPI and CPI.
  • [ ] I assigned the RAG color consistent with the numerical data.
  • [ ] I confirmed the reasons for deviation with reality.
  • [ ] The "what is needed from you" section of the report is clear.
  • [ ] I checked the integrity of the report against the watermelon effect.
Previous Next