Help

User Guide

Merge_me — User Guide

Merge_me is an interactive research instrument for learning IT integration decisions in post-merger scenarios. You work through realistic merger cases, make integration decisions across an analytical sequence, and get scored on the defensibility of your reasoning across eight competency dimensions.

v1 ships three healthcare scenarios; the primary is Premium Medical × Capital Clinics — two hospital groups merging their clinical and administrative IT. The others are RAKUS × Centrālā Laboratorija (a teaching hospital integrating a diagnostic lab) and Veselības centrs × Beauty clinic (a medical clinic merging with an aesthetic & wellness provider).

At a glance

A session has four parts and 9 stages of decisions:

Step 0 · The Deal — given context, not asked.
Phase 1 · System landscape (Stages 1–2) — map the integration landscape for this session's focus relationship.
Phase 2 · Per-pair evaluation (Stages 3–9) — score each canonical pair across cost, time, risk, stakeholders, users, goals, and criteria.
Report & Reveal — composite score + defensibility breakdown across 8 dimensions.

Each session focuses on one relationship type. When a session starts it is randomly assigned overlap, missing link, or gap, and the whole flow narrows to that one aspect — you map the system landscape for it and then evaluate only its integration options. Start a new session to work a different relationship type.

You can pause and resume any time. Decisions are saved per question.

Step 0 — The Deal

The session opens on a "deal sheet" showing:

Both companies and their industries
Merger type (concentric / horizontal / vertical / conglomerate)
Primary objective (cross-sell / cost synergy / vertical integration / etc.)
Two key capabilities — one to evolve, one to create
Time horizon (90 days / 6 months / 12 months / 18+ months)

This is the GIVEN context. You don't pick it — the scenario does. Read it, then click Begin Stage 1. The frame stays accessible from a collapsible "The deal" panel on every Phase-1 stage page.

Phase 1 — Mapping the integration landscape

Your session is assigned one relationship type — overlap, missing link, or gap. Phase 1 is that one relationship stage (Stage 1), which walks a five-question arc, followed by the constraints + triage stage (Stage 2).

Your session's relationship type

Overlap — both organisations do the same business activity. Example: Premium Medical and Capital Clinics both run patient billing & claims on separate systems.
Missing link — both have related work, but no connection runs between them yet. Example: Premium Medical's encounter history knows a patient's visits; Capital Clinics' device-data stream knows their vitals — but neither side can see the other.
Gap (no counterpart) — one organisation runs a process the other has no equivalent for. Example: Premium Medical files regulatory & accreditation reports; Capital Clinics has nothing equivalent. Pair entries may be one-sided.

The Step 0 deal sheet tells you which type this session focuses on.

The question arc (Stage 1)

Guess systems — build groups of Premium Medical ↔ Capital Clinics IT systems you think have this relationship. A hint chip tells you how many groups to aim for, and roughly how many systems per side. After you first submit, 2 systems per side are eliminated and shown crossed out — review them, then continue.
Process pairs — link processes that have this relationship between the two companies.
Refine systems — refine your groups. The canonical process answer is revealed above the form, and 4 systems per side are now struck-through and locked (the analysis has ruled them out).
Data pairs — link data entities that have this relationship.
Final systems mapping — confirm or change your groups. 6 systems per side are now eliminated, leaving 6 selectable. Each struck system carries a short reason inline, right under the row. The reason only names a dimension once you've actually engaged with it: at the first guess (nothing selected yet) it reads *does not support the merger goals*; after you've picked processes it reads *does not support a canonical process*; after you've picked data, *does not support canonical data*. (The canonical processes and data themselves are spelled out in the panels above the checklist.)

Justify on change (required, every step) — there is no separate "justify" step. On any of the system questions above (guess, refine, or final mapping), the moment you *change* a system you'd already committed to — including after the first elimination wave strikes systems on your guess — a required justification box appears right beneath the group you just changed, next to the checkbox. It names the exact change ("+ added Stripe", "− removed Looker") and you can't continue until you give a short reason. A written answer feeds dialogical fairness. Change nothing and no box appears — nothing is owed.

Competence meter

Phase 1 carries a competence meter, shown as a bar with a correct / total count (e.g. "5/7 systems"). The denominator is fixed at the canonical count — it's the target you're aiming for, so it never moves (a gap that needs 4 always reads "/4", it won't creep to "/5" when you add an extra). It is grouping-aware: a canonical system only counts when you place it in the *correct group*, not merely select it somewhere — each canonical group is matched to the player group it overlaps most, and only systems sitting in their matched group score. Three things pull the meter down: missing a canonical system, putting a right system in the wrong group (e.g. dumping everything into one big group), and over-selecting — extra/wrong picks are docked from the numerator (pick 5 when 4 are canonical and you'll see something like 3/4). So you can't reach 100% by selecting everything, nor by getting the systems right but the grouping wrong. If your final mapping scores below half, the round is lost and you start a fresh run.

Stage 2 — Constraints + Option triage

Two parts:

Context constraints — tick the regulatory, technical, and timing constraints the merger faces. Drives the cost / time / risk scoring downstream.
Per-pair option triage — for each canonical data pair, mark which integration options are realistic vs not feasible.

Phase 1 — Game mechanics

Hint chips

Each system-groups question shows an "Aim for ~N groups" hint chip sourced from the canonical pair count of the bucket's data layer.

System info icons

Every system in a checklist carries a small ⓘ icon next to its name. Hover or focus it for a one-line plain-language clarification of what that system does (e.g. *"Stripe — payment processing and card transactions"*). Clicking the icon never toggles the checkbox. Descriptions come from the scenario's systemsCatalog.

Running tally

Above each Q2 / Q3 / Q4 / Q5 question, a side panel surfaces your initial groups (or refined groups from Q3 once filled). Each pair is marked green if still viable or red + strike if ruled out by your subsequent picks.

Canonical process reveal (Q3)

Above the refine question, a panel reveals the scenario's authored canonical process pairs. Strike-through on systems below reflects systems not touched by any of these canonical processes. Educational moment: refine your systems against the right answer for processes.

Strikethrough + lock

On Q3 refine and Q5 final mapping, systems that aren't touched by the relevant prior picks (canonical processes on Q3; player processes + data on Q5) render as struck-through with a disabled checkbox. You can see the systems but can't toggle them. Lets you see your earlier guesses against the analytical conclusion without being able to erase the trail. Each struck row shows a short elimination reason inline, but it never names a dimension before you've selected it: the early guess wave reads *does not support the merger goals*; *does not support a canonical process* only appears once you've picked processes, and *does not support canonical data* only after you've picked data. The canonical processes and data themselves are named in the reveal panels above.

System info icons

Every system in a checklist — from the very first guess question onward — carries a small ⓘ icon next to its name. Hover or focus it for a one-line plain-language clarification of what that system does, so a participant unfamiliar with a tool can still pick confidently. Clicking it never toggles the checkbox.

Justify-on-change (required, on every system question)

This applies to all three system-groups questions — the initial guess, the refine, and the final mapping — not just the last one. Whenever a question carries a prior selection and you check or uncheck a system (for the guess, this means after you've submitted once and the first elimination wave has struck systems), a required justification box appears immediately, directly beneath the group whose checkbox you just toggled — so the ask sits right next to the change, not at the bottom of the form. It names the specific change — which system you added or removed — and the question can't be submitted until you give a short reason. A written answer feeds dialogical fairness. It only appears when there's an actual membership change to defend; making no change asks nothing.

Systems Review (per-stage results)

After each Phase-1 stage's Q5 submits, the results page shows a side-by-side comparison panel:

Left: your submitted groups, with each pick color-coded correct (green) or wrong (red + strike) against the canonical group set.
Right: the scenario's canonical groups, all rendered as correct (informational).

Same disabled-checkbox vocabulary as the form so the comparison reads directly.

Phase 2 — Per-pair evaluation

After Phase 1, the system surfaces the canonical pair set for your session's focus relationship — the data pairs the scenario evaluates against. Phase 2 analyses each pair across seven stages.

You can analyse pairs one at a time (focused, recommended) or roll through all of them in one sweep.

Stage 3 — Place your bets + Goal support

Top 3 prediction per pair, for short-term and long-term horizons.
Top goal: which merger goal you're optimising for.
Goal support: high / medium / low support for the chosen goal, per option per pair.

Stage 4 — Stakeholder weighting

Identify the top stakeholder + a skeptic stakeholder.
Rank options per stakeholder perspective, per pair.

Stage 5 — User group impact

Identify the top user group.
Rate user impact per option per pair.

Stage 6 — Cost (negative side)

Initial + ongoing cost ratings per option per pair. Context constraints from Stage 2 drive the canonical baseline.

Stage 7 — Time (negative side)

Initial + ongoing time ratings per option per pair.

Stage 8 — Risk (negative side)

Initial + ongoing risk ratings per option per pair.

Stage 9 — Gut pick + Criteria recheck

Two parts:

Gut pick: short-term and long-term winner per pair.
Criteria recheck: walk 7 decision criteria (cost, time, risk, user fit, stakeholder alignment, goal support, reversibility, regulatory fit), mark which apply, pick the best option per applicable criterion.

Option race chart

Every Phase-2 stage's results page shows an option race chart — a horse-race-style ranking of every integration option per pair. Updates after each question. Two lines per option:

Top line (filled): the canonical reading after the stage.
Background line (grey): your reading after the stage.

Your "place your bet" picks from Stage 3 carry 1st / 2nd / 3rd badges on their respective bars so you can see whether your initial calls hold up.

After Phase 1 — The system-groups overview

Between Phase 1 and Phase 2 you land on the system-groups overview page. It shows:

Per-pair status (analysed vs not started)
Player ID vs canonical pair comparison (did you identify each canonical pair?)
Quick-jump links into Phase 2 for each pair

Each pair can be analysed independently. Once every pair is analysed, the "Finish scenario" button finalises and computes the composite score.

Scoring

Each enabled stage contributes a weighted score to your composite. Decisions are scored across 8 dimensions:

Decision intent — clarity of the goal
Factual basis — accuracy on the facts of the case
Causal logic — coherence of cause and effect
Risk resolution — handling of uncertainty
Comparative rationality — quality of comparison across options
Practical viability — feasibility of the chosen path
Dialogical fairness — balance across stakeholder perspectives
Adaptivity & governance — reversibility and oversight

The two phases contribute to complementary halves of this profile. Phase 1 (perceiving the landscape) feeds factual basis (did you identify the right systems, processes and data), causal logic (do your system groups sit on the process→data chain), comparative rationality (did you tell the real systems from the distractors), and dialogical fairness (did you justify your grouping). Phase 2 (deciding) adds decision intent, risk resolution, practical viability, and adaptivity & governance.

An antibody scoring ≥ 8.0 in a phase banks a memory cell (a permanent, lasting unit of immunity). Each phase banks only its own antibodies, so your immune system updates in two waves: after Phase 1 (perception antibodies) and at completion (decision antibodies).

Decision-making immune system

Your hub shows your decision-making immune system — the eight dimensions reframed as antibodies, each defending against a challenge a reviewer would raise. (This replaces the old "competency organization.")

Immunocompetence (global level)

Your overall Immunocompetence level comes from accumulated IC points (composite_score × 100 per scenario):

L1: 0–300 · L2: 301–700 · L3: 701–1200 · L4: 1201–2000 · L5+: 2001+

Antibody tiers

Each antibody has a current tier (which can rise *and* fall) and a permanent immunisation badge (the highest tier ever reached):

Dormant — never challenged yet (asleep)
Sensitized — some immunity (low titer)
Resistant — solid immunity (mid titer)
Immune — full immunity (high titer); banks the immunisation badge at five memory cells

Immunity fades — boosters restore it

An antibody's titer (current strength) wanes over time if you don't re-train it — the current tier can drop to "waning" and a booster is recommended. Replaying is a booster: a faster, stronger response that rebounds above a first attempt. Stronger immunity (higher badge) fades more slowly. Your earned badges never expire, even when the titer wanes. (Fading is configurable and is part of the experiment.)

Each industry has its own immune system. v1 ships with one industry (healthcare).

Resuming a session

Every decision is saved per question. When you return to a scenario, you land on the first unanswered question. The "Welcome back" hub lists recent attempts and lets you resume any in-progress scenario or start a fresh one.

Privacy

Storage: EU region (GDPR-compliant — Frankfurt or Ireland).
Lawful basis: consent, with timestamp + version stored.
Retention: 24 months for raw data; aggregate data may be retained indefinitely.
Subject access / deletion: contact the researcher.
Pseudonymisation applied to the analysis dataset by default.

Glossary

Frame — the GIVEN context: industries, type, objective, capabilities, horizon.
Overlap — both companies do the same business activity.
Missing link — both companies have related work but no connection between them.
Gap — one company has something the other lacks entirely.
Canonical pair — the data pair the scenario is evaluating against (set by the scenario author).
System group — a cluster of Premium Medical + Capital Clinics IT systems that share a relationship (overlap / missing link / gap).
Antibody — a defensibility dimension in your immune system; it defends against one challenge a reviewer would raise.
Memory cell (HEP) — a permanent unit of immunity, banked when an antibody scores ≥ 8.0 in a phase.
Titer — an antibody's current strength (0–100); it wanes over time without re-training.
Booster — a re-played scenario that restores a waning antibody, rebounding faster and higher than a first attempt.
Immunocompetence — your global level, built from IC points (composite × 100 per scenario).
Composite — your weighted overall score per scenario.