Job search
Business Analyst Portfolio: Connect Requirements, Process Maps, and Acceptance Evidence
A complete fictional archive-intake case links sources, process branches, requirements and criteria, with executed local checks and pending handoff acceptance.
A business analyst portfolio should let a reviewer follow one change from its original need through the current process, proposed requirements, design choices and acceptance evidence. Connect each process branch to the requirement and result that explain it. A process map without requirement sources leaves decisions unexplained, while a completed test checklist without observed results leaves acceptance uncertain.
LandOffer publishes this guide and offers job-search tools. Compare recent tech roles with the analysis work your sample demonstrates. The archive, candidate and business records below are fictional. We executed an original AI-authored local traceability audit and rule model on October 21, 2026; no archive system, real stakeholder, physical collection, deployment or user acceptance test was accessed.

Choose one business change with a clear starting point
Choose a change that makes the relationships visible. Describe the problem, affected roles, existing handoffs and constraints before introducing a solution. Name the operation and decision owner so a reviewer can assess what “improving operations” meant in this case.
Separate the current process from the proposed process. Identify where an input originates, who evaluates it, what information moves and what happens when information is missing. Include an exception path. A diagram showing only the successful route hides the conditions that usually make analysis necessary.
Keep project scheduling outside the main evidence unless it explains a constraint. A delivery plan can show when work was intended to happen; it does not explain why a requirement exists or whether its outcome was accepted. Likewise, responsibilities copied from a job description do not show a completed analysis decision.
Name your contribution. You may have documented the process, elicited needs, reconciled conflicting requests or prepared acceptance criteria. Implementation and business approval might belong to other people. Describe those boundaries alongside the artifact rather than letting its polished appearance imply sole ownership.
Connect the need to the requirement and the proposed design
IIBA’s Business Analysis Standard discussion of requirements and designs distinguishes representations of needs from representations of possible solutions. It describes tracing relationships backward to business needs and forward to solution components. Use those relationships to make the origin and impact of a change inspectable.
A named field or queue is a design choice. The need it serves may be that an operator must distinguish an incomplete request from one ready for processing. Keep both statements so a reviewer can question the design without losing the underlying requirement. A different design could satisfy the same need.
Preserve requirement identifiers across the process note, criteria and result record. Make each identifier resolve to its statement and source so reviewers can inspect why the requirement exists. Linking R2 to every test can make a matrix look complete while obscuring which behavior each test actually examines.
Worked asset one: a complete archive-intake analysis packet
Fictional candidate Ari prepares an independent case for fictional Northbank Print Archive. The change concerns requests to reproduce an archive item. These stipulated records are teaching inputs, not interviews, actual collection rules or evidence of an implemented service.
The filled source register contains four items. S1, the intake coordinator’s invented process note, says requests sometimes arrive without an item reference or intended reproduction reason. S2, the fictional collections policy, requires specialist review for items marked fragile and prohibits self-review. S3, the dispatch clerk’s invented record, separates approval from the later issue receipt. S4, the service owner’s stipulated transition concern, requires staff to reconcile the physical handoff with its recorded receipt before replacing the existing process.
No actual frequency, processing delay or cost is attached to those statements. The proposed benefit is clearer request state and responsibility; it remains unmeasured. Ari’s scope covers intake and handoff evidence. It excludes copyright clearance, payment, scanning quality, inventory migration and public access to collection records.
The completed current-process map reads: requester sends a note → coordinator checks item and reason → missing information returns to requester → otherwise coordinator seeks review → approved request goes to dispatch → clerk records the issue receipt separately. The fragile-item and self-review rules exist in the fictional policy, but the current note does not establish how consistently they are applied.
The proposed map makes those branches explicit: capture required details → return incomplete request → check independent reviewer and item category → hold fragile items for a specialist → record review outcome → await receipt → record issued state after the receipt is present. The physical handoff remains a separate confirmation. Neither map is presented as a formally validated BPMN model or an observation of real staff behavior.
Here is the filled requirement and design record:
| Requirement and source | Outcome required in the fictional case | Proposed design and acceptance link |
|---|---|---|
| R1 from S1 | Incomplete requests remain distinguishable and return for information. | Item reference and reason fields; AC1 checks missing information. |
| R2 from S2 | A reviewer is independent; fragile items require a specialist. | Reviewer/category decision gate; AC2 checks independence and standard review, AC3 checks the fragile path. |
| R3 from S3 | Approval alone must not become an issued-item claim. | Separate awaiting-receipt and issued states; AC4 checks the receipt boundary. |
| R4 from S4 | Staff reconcile physical handoff and recorded receipt during transition. | Proposed handoff checklist; AC5 requires an intended-user walkthrough and matching receipt evidence. |
Ari’s finished story for the dispatcher is: “As the dispatch clerk, I need to distinguish an approved request awaiting a receipt from an issued request, so I can identify work still requiring a handoff record.” AC4 says the model must remain awaiting receipt when its other review conditions are satisfied but no receipt is present. An issued result requires the receipt condition as well as those review conditions.
The case compares two designs. A single approved checkbox is simpler, but merges review and issue. Separate states require more explanation but preserve S3’s distinction. Ari recommends retaining awaiting-receipt and issued as separate states, accepting the extra explanation needed to preserve S3’s handoff distinction. The business owner has not approved the design, and a real system would need to verify receipt identity, persistence, editing and authorization beyond this model.
Ari’s fictional contribution is the source register, two process maps, four requirements, five criteria and decision note. Our actual article artifacts were AI-authored. Readers should describe only analysis they personally performed, and label any fictional replacement material before presenting it as a work sample.
Write criteria that a result can actually answer
GOV.UK’s user-story guidance, last updated in May 2016 and still publicly available when checked, connects the actor, intended action and goal. It describes acceptance criteria as outcomes and encourages linking supporting evidence. This is service-delivery guidance, not a requirement that every employer use one story template.
For this case, AC1 means a missing item reference or reason returns needs-information. AC2 means independent eligible review is required and self-review is held. AC3 means an independent lead cannot bypass specialist review for a fragile item. AC4 distinguishes awaiting receipt from issued. AC5 concerns an actual physical handoff and intended-user confirmation, which an offline rule function cannot perform.
Record expected behavior before examining the implementation output. Otherwise a test can merely repeat whatever the implementation already does. Keep the criterion, input, expected result, observed result and decision together. Include a negative case that challenges a meaningful boundary.
Worked asset two: an executed traceability and acceptance ledger
We executed a new Python 3.12.14 exercise with four requirements, five criteria and six synthetic test records. It checks the declared relationships and runs a small request-state model. It does not operate a workflow application, validate an archive policy or verify a physical handoff.
The local result ledger is complete:
| Test and criterion | Synthetic input | Actual local result |
|---|---|---|
| T01, AC1 | Item reference absent; reason present. | Needs information. |
| T02, AC2 and AC4 | Standard item, independent lead, required details present, no receipt. | Awaiting receipt. |
| T03, AC3 | Fragile item, independent lead, required details present. | Held for specialist. |
| T04, AC3 and AC4 | Fragile item, independent specialist, details and receipt condition present. | Issued in the model. |
| T05, AC2 | Standard item, specialist reviewing their own request, receipt present. | Held for self-review. |
| T06, AC1 | Fragile item, specialist, reason absent. | Needs information. |
The executed ledger records six observed states matching six separately declared expected states, including the specialist hold and self-review denial. The source input’s SHA-256 hash remained unchanged. An issued result is a model label under stipulated inputs; no item moved and no real receipt was verified. These six cases are not exhaustive tests of arbitrary inputs, permissions or concurrent changes.
The traceability check found no broken references in the baseline. It separately identified AC5 as having no executed test. This is the intended result: physical handoff evidence remains pending. Keep structural traceability, observed behavior and stakeholder acceptance as separate result fields; AC5 remains pending despite the consistent baseline.
Two altered records show how the audit handles contradictions. Changing R2’s source from S2 to nonexistent S9 produces an unknown-source finding. Assigning AC3 to R1 while the requirement record still places it under R2 produces wrong-owner and unlinked-criterion findings. The audit rejects both altered records rather than reporting a complete matrix from identifier counts alone.
A deliberately overbroad rule variant ignores the fragile category. For T03 it returns awaiting receipt, whereas the declared criterion requires a specialist hold. That counterexample rejects the variant. It does not prove the retained implementation correct for every possible policy or input.
The completed disposition is: retain the corrected local model and consistent trace record; hold any real release decision. AC5, intended-user checks, operational implementation and real owner acceptance remain pending. The result is a finished analysis sample with an explicit unresolved boundary, not a completed production change.

Use traceability to explain a change request
Suppose the fictional owner revises S2. The executed relationship query identifies R2, AC2 and AC3, and tests T02 through T05 as affected. That result gives a reviewer a concrete starting point for impact analysis. It does not say the policy was actually changed or the identified list covers every real dependency.
Ari’s filled change note is: “Revisit the independent-review and fragile-item criteria, review the affected local tests, and reconcile the decision gate with the proposed map. Keep R1’s completeness rule and R3’s receipt boundary visible while checking whether the revised policy affects them indirectly.” This explains the proposed review work without inventing a stakeholder meeting or approval.
IIBA’s public glossary distinguishes verification of artifact quality, validation of intended use and user acceptance assessment against identified criteria. Keep those outcomes separate in your portfolio. The local consistency check concerns artifact relationships; the model tests concern finite behavior; neither establishes that real users accept the solution.
Package a sample a hiring reviewer can question
Put a short index before the artifacts: context, sources, current and proposed maps, requirements, criteria, results, change note and unresolved decisions. Preserve one version across them. If the map changes after testing, identify the revision that generated the results rather than letting an older passing log endorse a newer design.
For confidential work, share only material you have permission to disclose. A newly invented example can explain your method, but its sources and outcomes must remain fictional. It cannot inherit adoption, time savings or stakeholder agreement from an undisclosed project.
A defensible application summary names the stage reached: “Prepared a synthetic intake-change analysis linking process branches, requirement sources and acceptance criteria; executed finite model and traceability checks and retained a pending physical-handoff criterion.” Use equivalent wording only for work you actually performed and can explain.
Open one source record, follow its requirement and criterion, and inspect the corresponding result before attaching the case. Then use LandOffer’s current listings to compare the role’s elicitation, process analysis, requirement management and acceptance responsibilities with the evidence your sample contains.
Sources and Further Reading
- GOV.UK: Writing user stories, last updated May 23, 2016; current public guidance checked October 21, 2026.
- IIBA: Understanding Requirements and Designs, public text retrieved October 21, 2026; ordinary browser navigation redirected to login, and no account was accessed.
- IIBA: BABOK Guide glossary, public definitions checked October 21, 2026.


