SELF-GUIDED BA PRACTICE Packages available · Access after confirmed paymentSee packages ↗

Fundamentals

Business Case Template: A Worked Example for Junior Business Analysts

Fundamentals · Published · Updated · 12 min read · By BA Mentorship Editorial (AI-written)

Part of our guide to business analyst roles and responsibilities.

What a business case is and why it exists

A business case is a decision document. It helps a sponsor and a decision group answer one question: is this change worth the time, money and disruption, and which way of doing it is best? It is usually written before a project gets full funding. It is revisited later to check whether the promised benefits arrived.

A good business case is short, honest and easy to compare. It does not sell one answer. It shows the problem, the realistic options, what each option costs, what each option returns and what could go wrong. The decision makers then choose with their eyes open.

The standard structure

Organisations use different templates, but most contain the same building blocks. Use your company's template if it has one. If not, this structure works well for small and medium changes.

TEXT

BUSINESS CASE: [Project name]
Version | Date | Author | Sponsor

1. Executive summary
   Problem, recommended option, key benefits, decision requested
2. Background and problem statement
3. Objectives (measurable, with target dates and their source)
4. Scope (in scope and out of scope)
5. Options considered
   Option A: Do nothing (baseline)
   Option B: [low-effort option]
   Option C: [larger option]
6. Costs per option (one-off and ongoing, with source and status)
7. Benefits per option (measure, baseline, target, source, owner)
8. Risks per option
9. Assumptions, constraints and dependencies
10. Recommendation and reasoning
11. Next steps and decision requested

The order matters. Someone who only reads the executive summary should still know the problem, the recommendation and the decision being asked for. Readers who want detail can jump straight to the options, costs and assumptions.

How a BA contributes versus the sponsor

The sponsor owns the business case. They are accountable for the decision, they present it to the steering group or budget holder, and later they answer for whether the benefits were delivered. The BA is usually a contributor, and often the main author of the analysis sections.

As a junior BA you will rarely write a whole business case alone. You are more likely to gather evidence, draft the problem statement, build the options table and keep the assumptions log tidy. That work matters, because weak evidence and hidden assumptions are where most business cases fail. For the wider picture of the role, see what a business analyst does.

Who does what in a business case

AspectProject sponsorBusiness analyst
OwnershipAccountable for the case and the final decisionSupports the sponsor and writes key analysis sections
Problem statementConfirms the problem matters to the businessGathers evidence and drafts a clear, solution-free statement
OptionsSets limits such as deadline and funding rangeIdentifies and analyses options, including do nothing
Costs and benefitsAgrees benefit targets and who will deliver themCollects estimates from owners and records each source
Risks and assumptionsDecides what level of risk is acceptableLogs assumptions and risks, and tests the biggest ones
RecommendationPresents it and asks for approvalDrafts the reasoning that links evidence to the choice
A typical split of responsibilities. The exact split varies by organisation, so confirm it with your sponsor.

The project manager, if there is one, usually supplies the delivery plan and effort estimates. The difference between the two roles is explained in Business Analyst vs Project Manager. Finance often checks cost figures and any return calculations.

How to build a business case, step by step

These steps follow the template above. In real projects they overlap: new evidence in step five may send you back to step four. That is normal.

  1. Agree the purpose and audience. Ask the sponsor who will decide, what they need to see and when. A quick stakeholder analysis tells you whose views must appear in the case.
  2. Define the problem. Gather evidence such as volumes, handling times, error counts and complaints. Write a short problem statement that says what is wrong, who it affects and the impact, without naming a solution.
  3. Set measurable objectives. Turn the problem into targets the business can check later. Write down where each target comes from, for example a capacity gap or a service commitment.
  4. Identify options. Brainstorm widely, then shortlist two or three realistic options plus do nothing. Include at least one low-effort option.
  5. Estimate costs and benefits. Ask the people who own each number. Where you can, test performance estimates with a small trial. Record the source and confidence of every figure, and log every assumption.
  6. Compare and recommend. Put the options side by side in the same format. Recommend only when the key cost lines are in. If some are still pending, mark the recommendation as provisional.
  7. Review and hand over. Walk the sponsor through the draft, fix gaps, keep a version history and store it where the team can find it.
Building a business case, step by step7 steps in order: 1. Agree purpose and audience: Ask the sponsor who decides, what they need to see and when the decision is due. 2. Define the problem: Write a problem statement backed by evidence such as volumes, times and complaints. 3. Set objectives: Turn the problem into measurable targets and record where each target comes from. 4. Identify options: List realistic options, always including do nothing, then shortlist two or three. 5. Estimate costs and benefits: Collect estimates from owners, test with small trials and log every assumption. 6. Compare and recommend: Compare options side by side once key costs are in, then draft the recommendation. 7. Review and hand over: Walk the sponsor through the draft, fix gaps and keep a version history.1Agree purpose and audienceAsk the sponsor who decides, what theyneed to see and when the decision is due.2Define the problemWrite a problem statement backed byevidence such as volumes, times andcomplaints.3Set objectivesTurn the problem into measurable targetsand record where each target comes from.4Identify optionsList realistic options, always includingdo nothing, then shortlist two or three.5Estimate costs and benefitsCollect estimates from owners, test withsmall trials and log every assumption.6Compare and recommendCompare options side by side once keycosts are in, then draft therecommendation.7Review and hand overWalk the sponsor through the draft, fixgaps and keep a version history.
Seven steps from the first conversation to a draft the sponsor can own. Steps often overlap and loop back as evidence arrives.
Diagram in words
  • 1. Agree purpose and audience: Ask the sponsor who decides, what they need to see and when the decision is due.
  • 2. Define the problem: Write a problem statement backed by evidence such as volumes, times and complaints.
  • 3. Set objectives: Turn the problem into measurable targets and record where each target comes from.
  • 4. Identify options: List realistic options, always including do nothing, then shortlist two or three.
  • 5. Estimate costs and benefits: Collect estimates from owners, test with small trials and log every assumption.
  • 6. Compare and recommend: Compare options side by side once key costs are in, then draft the recommendation.
  • 7. Review and hand over: Walk the sponsor through the draft, fix gaps and keep a version history.

Worked example: a refund process business case

Marigold Home Supplies is a fictional online retailer. Its customer service team handles refund requests by email. Agents copy each request into a spreadsheet, look up the order in a separate system, then email finance to pay. Helena, the head of customer service, is the sponsor. Sunil is the operations lead. You support Nadia, the senior BA.

Problem and objective, with sources

The weekly ticket report shows about 240 refund requests a week. You observed 20 of them and timed each one: they took 15 minutes on average. So refunds use about 60 agent hours a week (240 requests times 15 minutes). Sunil's rota records show the team works about 20 hours of overtime a week, mostly on refunds.

The objective comes from that overtime figure, not from a guess. To free 20 hours a week at the same volume, refunds must take about 40 hours a week. That is 10 minutes per request (240 times 10 minutes). Objective agreed with Helena: 10 minutes per request within six months of go-live, with no regular refund overtime.

Options analysis

  • Option A, do nothing. Handling stays at about 60 hours a week, and grows if volumes grow. No project cost, but the 20 hours of overtime and the delay complaints continue.
  • Option B, improve the current process. A standard web form, email templates and fixed spreadsheet columns. Source: two agents trialled this for one week on 20 real requests and averaged 12 minutes. At 240 a week that is about 48 hours, saving about 12.
  • Option C, refund module in the existing helpdesk tool. A form linked to order lookup, with a queue for finance. Source: Riya and one agent processed 10 sample requests in the vendor's trial environment and averaged 10 minutes. That gives about 40 hours a week, saving about 20.

Costs and the comparison

Option A is not a straw man. It is the baseline every other option is measured against. At first the cost lines were empty. You did not fill them. They were marked pending with owners: the licence quote with Daniel, the integration estimate with Mei, the Option B effort with Sunil. While they were pending, the draft said recommendation to follow when costs are confirmed. Once the figures arrived, finance valued agent time at its internal rate of $20 per hour over 50 working weeks a year, and the options could be compared.

Marigold refund options compared

AspectOption A: do nothingOption B: improve processOption C: helpdesk module
Agent hours per weekAbout 60About 48 (one-week trial, 20 requests)About 40 (vendor trial, 10 requests)
Hours saved per weekNone, this is the baselineAbout 12About 20
One-off cost$0$1,000 (Sunil's effort estimate, costed by finance)$8,000 (Mei's integration estimate, costed by finance)
Ongoing cost per year$0$0$4,000 licence (vendor quote via Daniel)
Value of hours saved per year$0$12,000 (600 hours)$20,000 (1,000 hours)
Net value, year one$0$11,000$8,000
Net value, year two$0$12,000$16,000
Meets the objective?No, 20 hours of overtime continueNo, about 8 hours of overtime remainYes, if trial times hold in live use
Fictional figures. Value of hours saved uses finance's rate of $20 per hour over 50 working weeks. Net value is value of hours saved minus costs for that year.

Check the sums. Option B: 12 hours times 50 weeks is 600 hours, worth $12,000; minus $1,000 gives $11,000 in year one. Option C: 20 hours times 50 weeks is 1,000 hours, worth $20,000; minus $8,000 and $4,000 gives $8,000 in year one, then $16,000 in year two. Over two years, B gives $23,000 and C gives $24,000.

So the recommendation is honest about the trade-off. Option B is better in year one and the two-year gap is small. The main reason for Option C is that it is the only option that meets the objective and removes the regular overtime. The case also logs two assumptions: volume stays near 240 a week, and the trial-environment time of 10 minutes holds in live use.

Summary page of the Marigold business case

Marigold refund business case: summary

Problem: 240 refund requests a week at 15 minutes each use about 60 agent hours, plus about 20 hours of overtime.1

Objective: 10 minutes per request within six months of go-live, removing regular refund overtime.2

Option A, do nothing: about 60 hours a week continues. No project cost.3

Option B: about 48 hours a week (one-week trial). One-off cost $1,000.

Option C: about 40 hours a week (vendor trial). One-off $8,000, licence $4,000 a year.

Sources: vendor quote (Daniel), integration estimate (Mei), effort estimate (Sunil), hourly rate (finance).4

Assumptions: volume stays near 240 a week and trial times hold in live use.5

Recommendation: Option C, with Option B templates live now as a quick win.6

  1. Evidence, not opinion. Volume comes from the ticket report, time from 20 observed cases and overtime from rota records. 240 times 15 minutes gives about 60 hours, the baseline.
  2. Objective with a source. The 10 minute target is the time per request needed to free the 20 hours of overtime: 240 times 10 minutes is 40 hours.
  3. Do nothing as baseline. Option A shows the cost of no change. Option B saves about 12 hours a week and Option C about 20 against it.
  4. Every figure has an owner. The BA collected and checked the inputs. Costs came from their owners and finance applied its own rate. Nothing was guessed.
  5. Small samples, stated openly. Both trials were small, so the case says so. If volume or live handling times differ, the hours saved and the value change.
  6. Recommendation follows. B is better in year one, but only C meets the objective. By the end of year two C is also ahead by $1,000 in net value.
Each line maps to a template section. The notes explain why it is written that way.

Defending the recommendation

Marigold Home Supplies, a fictional retailer: the BA team reviews the completed refund business case with the sponsor before the steering group.

  1. Helena Project sponsor

    Option B gives more net value in year one. Why are we recommending Option C?

  2. You The BA learner

    Option B leaves about 8 hours of refund overtime a week. Only Option C reaches 10 minutes per request, which is what removes the 20 hours.

  3. Nadia Business analyst

    By the end of year two, Option C is also about $1,000 ahead on finance's rate. That gap is small, so the objective is the main reason, and the case says so.

  4. Helena Project sponsor

    Why keep do nothing in the table at all?

  5. You The BA learner

    It is the baseline. Without the 60 hours a week, nobody could see that B saves 12 hours and C saves 20.

  6. Helena Project sponsor

    Fine. The Option C time came from a trial environment, so flag that as a risk. I will take it to the steering group. The decision is mine.

The people and the company in this scene are fictional.

Once the cost figures arrived, the BA explains the trade-off and the baseline to the sponsor.

Outcome in the story: the steering group approved Option C. It asked for the Option B email templates to go live straight away as a quick win, and for live handling time to be checked three months after go-live.

Writing costs, benefits and assumptions without inventing numbers

This is where junior BAs feel the most pressure. Someone asks for a rough number and it is tempting to type one in. Resist. Once a figure appears in a business case, people repeat it as fact, and the person who typed it owns the problem.

Costs

List cost lines for each option, then get figures from the people who own them. Separate one-off costs from ongoing costs, because ongoing costs are easy to forget.

  • One-off: build or configuration effort, upfront licences, data migration, training, temporary extra staff around go-live.
  • Ongoing: subscriptions, support, maintenance, hosting and the staff time needed to run the new process.
  • For each line: the source (vendor quote, IT estimate, finance), the owner, the status (confirmed, estimate or pending) and the confidence level.

Benefits

Tangible benefits can be measured, such as hours saved, errors avoided or faster replies. Intangible benefits, such as better staff morale, are real but harder to measure, so describe them honestly and do not dress them up as figures. For each tangible benefit, write the measure, the current baseline, the target, the source of the estimate, the owner and when it will be checked.

A small trial is often the best source for a performance estimate. Timing a handful of real or sample cases is better than a vendor brochure or a hopeful guess. State the sample size so readers know how much weight to give it.

Finance teams often turn costs and benefits into return or payback calculations. Your job is to give them clean, sourced inputs, not to replace their methods.

Assumptions, constraints and dependencies

An assumption is something you believe is true but have not proved, such as stable volumes. A constraint is a fixed limit, such as a deadline. A dependency is something you need from elsewhere, such as another team's system. Log each one, note what happens to the case if it proves false, and test the biggest ones first. If the real question is whether an option is possible at all, a feasibility study may need to come first.

Common mistakes junior BAs make on business cases

  • Skipping do nothing. Without a baseline, nobody can see what the change actually buys.
  • Selling one option. If the options table is written so one option must win, reviewers notice and trust drops.
  • Guessing numbers. A placeholder with an owner is better than a confident guess that later proves wrong.
  • Targets with no source. An objective should come from a real need, such as a capacity gap, not from a number that sounds good.
  • Recommending before costs are known. If key cost lines are pending, label the recommendation provisional or wait.
  • Hiding assumptions. Every estimate rests on assumptions. If they are not written down, nobody can test them.
  • Benefits nobody owns. A benefit with no owner and no measure is unlikely to be checked after go-live.
  • Forgetting ongoing costs. Licences, support and training continue long after go-live.

Most of these mistakes come from rushing. If the sponsor needs the case quickly, agree which sections can be lighter, rather than cutting corners on evidence.

Business case review checklist

Use this list as a final check before the draft goes to the sponsor. If any item fails, fix it or flag it clearly as open.

  • The problem statement is backed by evidence, not opinion, and names no solution.
  • Objectives are measurable, have a target date and state where the target comes from.
  • Do nothing is included and described honestly as the baseline.
  • Each option shows costs, benefits, risks and key assumptions in the same format.
  • Every cost and benefit names a source, an owner and a status.
  • Missing figures are marked pending with an owner, not guessed.
  • The recommendation is based on a complete comparison, or is clearly marked provisional.
  • The sponsor has reviewed the draft and agreed to own it.

Frequently asked questions

How long should a business case be?

It depends on the size of the change and your organisation's rules. A small process change may need only a few pages, while a large programme may need a full document with appendices. Use your company's template if one exists, and keep the executive summary short enough to read in a few minutes.

Is a business case the same as a feasibility study?

No. A feasibility study checks whether options are possible, for example technically, legally or operationally. A business case uses that information to argue which option is worth funding. On larger projects the feasibility work often comes first and feeds the options section.

What should I do if nobody can give me cost figures?

Do not fill the gap yourself. List the cost line, mark it as pending, and record who will provide the figure and by when. If an owner can only give a range, write the range and the reason for the uncertainty. Tell the sponsor early if missing figures could delay the decision.

Can the recommended option be worse on cost than another option?

Yes. The best option is the one that meets the objectives at an acceptable cost and risk, not always the cheapest. Say the trade-off openly in the recommendation so decision makers can weigh it themselves.

Does the business case change after approval?

It should be revisited at key points, for example when scope, costs or risks change significantly. After go-live, the benefits listed in the case are checked against actual results. That is why measurable benefits with clear owners matter so much.

This article was written by an AI system and published automatically after automated accuracy, safety and quality checks. All companies, people and numbers in the examples are fictional. BA Mentorship does not issue professional certifications and cannot guarantee any job or interview outcome. Spot an error? Email our support team and we will check and correct it.

Key terms in this article

Browse the full glossary

CHECK BEFORE CONTINUING

Keep your work safe