Skip to content

Lab 23 — Agile versus Predictive Planning

At a glance

Learning objectives

By the end of this lab you should be able to:

  1. Compare predictive and adaptive delivery on the dimensions that actually differ, rather than on slogans.
  2. Quantify the cost of absorbing change under each approach.
  3. Recommend an approach for a specific project, and state the conditions that would reverse the recommendation.
  4. Recognise where a hybrid shape is the honest answer instead of a compromise.

Scenario

The E-Commerce Platform from Lab 14 is at a decision point. The retailer's owner must choose how the work will be planned and controlled for the next four sprints. Her board has two proposals.

Proposal P — Predictive. Baseline a fixed scope of 60 story points covering the five highest-priority features. Fix the date at four sprints (eight weeks) and fix the budget. Any change enters through change control, is assessed for impact, and the baseline is revised only with the owner's approval.

Proposal A — Adaptive. Maintain a single ordered backlog of all thirteen items (100 points). Plan each sprint from an updated order. Same date, same budget. A change is admitted by removing something of comparable size from the current sprint or from the backlog.

Delivery facts apply to both:

Item Value
Measured velocity 18 points per sprint
Sprints remaining 4
Capacity available about 68 points
Team 3 developers, 1 tester, 1 designer
Date commitment Winter season start, eight weeks away
Budget Fixed

Change data from the last comparable project, invented for teaching:

# Change Points When it arrived
C1 Live chat support on product pages 8 After sprint 1 of 6
C2 Replace the payment provider 13 After sprint 2 of 6
C3 Add consent capture for marketing preferences 5 After sprint 3 of 6

Observed effects, from the same data:

  • Under a predictive baseline, each change costs about 1.4 × its size in consumed capacity, because re-planning, regression testing, and integration rework are additional to the change itself.
  • Under an adaptive order, each change consumes 1.0 × its size, and the cost is paid by deferring items of comparable size rather than by exceeding capacity.
  • The retailer's own history shows roughly three material changes per two months of delivery, and that rate has been stable for a year.
  • The winter season start is a commercial commitment; missing it costs the retailer its most valuable trading period.

Team and stakeholder position:

  • Developers are indifferent between the two, but the tester prefers adaptive because regression testing is smaller per increment.
  • The business owner wants the date held and says the only items she insists on are guest checkout, payment retry, and order tracking.
  • The marketing manager supports Proposal P because a fixed baseline makes her campaign planning simpler.
  • A supplier promotion arriving in six weeks will require gift cards (13 points) if it is accepted.

Your role

You are the delivery consultant advising the retailer's owner. You are not selling a methodology: you must recommend what will actually work for this business, and be candid about what each option will feel like when it goes wrong.

Project information and assumptions

The scenario is part of the example-project library — read the full description there if you want the wider context.

  • Capacity is about 68 points across four sprints in both proposals.
  • Change effects (1.4 × vs 1.0 ×) are given; use them as stated and state their source in your answer (they are assumed for teaching, and you must say so).
  • Assume the change rate of three material changes per two months continues.
  • Assume no change can be refused outright: the business will either accept it or accept the consequence of not accepting it.
  • Do not treat either proposal as automatically superior. Both have been used successfully in real organisations, and for different reasons.

Tasks

Task 1 — Establish the baseline arithmetic

For both proposals, state:

  1. The scope committed to at the start, in points.
  2. The capacity available, in points.
  3. The buffer between them, and what that buffer is for.

Then answer: Proposal P baselines 60 points against 68 available. What is the 8-point difference for, and what happens if it is consumed by change?

Task 2 — Quantify the cost of change

Using the change data and the given multipliers:

# Change Points Predictive consumption (1.4×) Adaptive consumption (1.0×)
C1 Live chat 8
C2 Payment provider 13
C3 Consent capture 5
Total 26

Then calculate:

  1. The total capacity consumed under each approach.
  2. Each total as a fraction of the 68-point capacity.
  3. Under Proposal P, how many sprints of capacity the change consumes, and what that does to the date or to the scope.
  4. Under Proposal A, how many points must be deferred, and name the two items you would defer first.

Task 3 — State the consequence each proposal produces

For each proposal, write a short paragraph describing what actually happens to the retailer — not to the methodology — when the three changes arrive.

Your paragraphs must each include:

  1. What is delivered by the winter season start.
  2. What is not delivered.
  3. Who inside the retailer is most affected, and how they will react.
  4. The single risk that would most damage the outcome.

Task 4 — Compare on eight dimensions

Build a comparison table across these dimensions, with a rating and a one-line justification for each:

Dimension Predictive (P) Adaptive (A)
Scope clarity at the start
Date confidence
Cost confidence
Cost of absorbing change
Risk early-warning
Measuring progress
Stakeholder expectation setting
Suitability for a fixed-price contract

Then answer: on which of the eight dimensions does the answer not favour either approach in general, but depend on this project's specifics? Name them and state the specifics.

Task 5 — Handle the gift cards question

The supplier promotion in six weeks would require gift cards (13 points, dependent on guest checkout).

  1. State what Proposal P can do about it, and the consequence.
  2. State what Proposal A can do about it, and the consequence.
  3. Decide whether the promotion is worth admitting under either proposal, and justify with numbers.

Task 6 — Recommend

Write a decision memo of no more than 300 words for the retailer's owner, containing:

  1. The approach you recommend, in one sentence.
  2. The three strongest reasons, each with a number from your own work.
  3. What she will have to accept as a consequence of your recommendation.
  4. Three conditions that would reverse your recommendation, each stated concretely.

Then, separately, answer: is there a hybrid shape that would serve this project better than either pure proposal? If there is, describe it and say what makes it better than a compromise. If there is not, say so and explain why the hybrid would be worse.

Task 7 — Design the reporting for whichever you chose

Whichever proposal you recommended, the owner needs to know how she will learn that it is going wrong.

  1. Name the four indicators you would report, and the cadence.
  2. State the threshold on each at which she should worry.
  3. State one indicator you would deliberately not report, and why reporting it would cause harm.

Deliverable

  1. The baseline arithmetic for both proposals, with the buffer explained.
  2. The cost-of-change table with the four calculations.
  3. Two consequence paragraphs, one per proposal.
  4. The eight-dimension comparison table and the "depends on specifics" answer.
  5. The gift cards analysis.
  6. A 300-word decision memo with three reversal conditions, plus the hybrid verdict.
  7. The reporting design with thresholds and the deliberately omitted indicator.

Hints and guidance

  • The 1.4× multiplier is the crux. Under a fixed baseline, change is not just more work; it is re-planning, regression, and rework.
  • Both proposals hold the date. The difference is what is sacrificed when the date is held. Under P it is the date or the budget; under A it is scope, chosen deliberately.
  • Deferring is not failing. Under A, the deferred items are the answer to "what changed?", and being able to name them is a sign the approach is working.
  • Slogans are not arguments. "Agile is more responsive" and "predictive is more disciplined" are both true in some contexts and useless in yours.
  • Say where your numbers came from. The change rate and multipliers are assumptions for teaching; stating that is part of professional practice.

Self-check criteria

Check How to verify it
Change totals correct Predictive 36.4 points; adaptive 26 points
Fractions of capacity correct 36.4 ÷ 68 ≈ 54%; 26 ÷ 68 ≈ 38%
Consequence paragraphs name who is affected Not "the business" but specific roles
All eight dimensions rated No blank cells; each with a justification
Reversal conditions are concrete Each names an observable condition, not "if things change"
Memo uses at least three numbers From your own tables
Hybrid verdict is argued either way A reasoned answer, not a hedge
Assumptions labelled as assumptions The multipliers and change rate are stated as given for teaching

Reflection

  1. The predictive approach consumes 26 points of change as 36.4 points of capacity. If the change rate halved, would your recommendation change? Show the arithmetic.
  2. The marketing manager prefers Proposal P because it is simpler to plan campaigns around. Is that a legitimate argument? How would you handle it — with evidence, or with a concession?
  3. Which risk does Proposal A make worse rather than better, and how would you mitigate it inside an adaptive approach?
  4. If the retailer were contracting this work to an external firm on a fixed price, which proposal would the contract force, and what would you warn the retailer about as a result?
Check your numbers

Cost of change

Change Points Predictive (1.4×) Adaptive (1.0×)
C1 8 11.2 8
C2 13 18.2 13
C3 5 7.0 5
Total 26 36.4 26

As a fraction of capacity - Predictive: 36.4 ÷ 68 ≈ 54% of all available capacity consumed by three changes. - Adaptive: 26 ÷ 68 ≈ 38%, paid by deferring 26 points of lower-value work.

Consequence under Proposal P: the 8-point buffer is consumed by C3 alone, and the remaining 28.4 points of change force either a two-sprint slip or the removal of approximately 28 points of baselined scope — more than the three features the owner says she insists on combined.

Consequence under Proposal A: the date holds; roughly 26 points of lower-value work is deferred — for example the recommendation strip (13 points) and loyalty points (13 points) — and the owner sees exactly what she traded away.

Full working is in the Advanced labs answer key.

Extension challenge

Repeat the analysis for a project where the change rate is near zero and the requirements are fixed by regulation — for example a statutory reporting change with a legislated date. Show how the arithmetic changes, then answer two questions honestly: (a) does predictive planning win on the numbers, or does it merely win on the contract? and (b) what does an adaptive team do about a legislated date, given that deferring scope is not available when the scope is the regulation?