APEX Portfolio Forecasting — Johnson & Johnson

Designed APEX, a scenario-planning platform for Johnson & Johnson's forecasting and strategy teams — replacing fragmented Excel workflows with a versioned, interactive environment for evaluating portfolio decisions across Atopic Dermatitis and beyond.

AICommercialHealthcareEnterpriseWeb

Guidehouse: AI Studio – J&J (Health Commercial) · 2026

Overview

A clear entry point for a complex system

APEX replaces Excel-based pharmaceutical forecasting with a versioned, interactive workspace : scenario planning that can be audited, compared, and defended in an executive review. Built as a one-week RFP response for J&J's commercial strategy team, covering US Atopic Dermatitis with a clear expansion roadmap.

Pick an indication — AD, Asthma, or COPD
Navigate to Market Map or Portfolio Forecast Builder
Load or branch a scenario from the shared library
Expansion markets (UK, FR, DE, ES) visible as coming soon from day one

New scenarios branch from a shared baseline. Existing ones load directly into the builder.

APEX home screen showing indication tiles for Atopic Dermatitis, Asthma, and COPD with a scenario library table below

Home — indication selection, navigation, and scenario library

The Problem

Excel forecasting doesn't survive a methodology review

Scenarios lived in separate Excel files — no single source of truth
Assumptions duplicated across versions with no methodology tracking
Prior outputs couldn't be reconstructed when methods changed
No audit trail: by the time a decision reached an executive review, nobody could trace the logic behind the numbers

The RFP framed the solution: build a single versioned environment where teams can test assumptions, compare scenarios, and export results that can be defended six months later.

Role: Lead Designer + acting PM
Timeline: 1 week, RFP to pitch
Scope: End-to-end prototype, MVP scoping to pitch
Team: Health SMEs, pharma strategy leads, J&J stakeholders
Indication: US Atopic Dermatitis (MVP) → Respiratory (roadmap)

HowIapproachedit

This was a one-week RFP response with no dedicated PM, so I took on that role: collecting requirements, organizing the team, and running a 5-hour brainstorming session to align on direction. The team leads were new to Guidehouse and unfamiliar with its capabilities, so I helped orient the engagement while iterating on the design in parallel. Engineering hadn't started, and every decision had to be defensible enough to win the engagement and detailed enough to hand off directly to devs.

Domain grounding

Mapped the AD market structure before designing any screens
Covered epidemiology, competitive landscape, payer dynamics, and site of care
Built enough domain fluency to make defensible design decisions, not just wireframes

MVP scoping

Scoped to US Atopic Dermatitis: Market Landscape Dashboard, TPP Builder, Portfolio Forecast Builder
Small expert user group → depth and fast feedback over broad coverage

Scenario system design

Designed the scenario library as the central organizing structure
Named snapshots, editable market inputs, engine selection, compare mode, export flows
Every screen anchored to a question a strategy leader needs before an executive review

Governance layer

Built assumption versioning, sandbox vs shared workspace, and publishing controls from the start
Reproducibility shaped how the scenario object was structured, not added as an afterthought

APEX pitch packaging

Renamed from FutureCast to APEX and refined the prototype for executive storytelling
Simplified charts, added Price Optimizer, updated source-entry patterns across input forms
Prepared design for J&J stakeholder review

Portfolio Forecast Builder

The main workspace

The builder is where forecasting happens: three-panel layout, live chart updates, no separate recalculation step.

Left rail — market inputs, editable inline
Main panel — forecast chart, Volume or Revenue metric
Right panel — scenario list, Product or Market view toggle
Including or removing a product updates the chart and scenario immediately

Portfolio Forecast Builder showing patient volume forecast for 4 products across 2026–2045 with base case loaded

Patient Volume Forecast — 4 products, Base Case loaded

Builder states

Revenue view — 1 product, market view

4 products, two scenarios checked — Compare Selected active

Scenario Compare

Two scenarios, one view

Compare mode loads when users select scenarios from the library and hit Compare. The chart switches to a multi-line overlay : Base Case renders darker, Best Case medium, Worst Case lighter. The layout stays inside the builder. Exit Compare returns to the last active state without losing the current product selection or view mode.

Revenue — Base vs Best Case overlay

Volume — Base vs Best Case overlay

Market Intelligence Layer

The data behind the assumptions

The Market Map surfaces the epidemiological and competitive context behind every assumption. Four views cover the full picture:

Total Prevalence — overall population trajectory
Subpopulation — Moderate, Severe, Pediatric, Refractory segments
Payer Mix — Commercial, Medicare, Medicaid, Other breakdown
Product Treatment Rates — competitor landscape by drug

The Market Insights panel on the right tracks the current selection with market size, growth rate, and key drivers.

Total Prevalence — Overall population view

$Market Map with subpopulation toggle active showing Moderate, Severe, Pediatric, Refractory segments$

Total Prevalence — Subpopulation breakdown

Patient Distributions — Payer mix

Product Treatment Rates — Competitor landscape

Assumption Editing

Every input is explicit, every source is linked

Six modal forms cover the full market input stack. Each follows the same structure: current value, future value (with a target year), and a source field backed by a dropdown with an add-new control.

Prevalence — US population base, AD prevalence rate, growth method
Subpopulation — adult and pediatric segment distribution, current and future
Diagnosis Rate — adult and pediatric diagnosed fraction with growth rate
Overall Treatment Rate — treated fraction across the diagnosed population
Payer Mix — commercial, Medicare, Medicaid, cash distribution
Site of Care Mix — outpatient vs retail channel split

Modal patternSame structure repeats across all 6 forms

Edit Prevalence Inputs modal with US Population, AD Prevalence %, Growth Rate, Method, and Source fields

Edit Prevalence — population base, rate, growth method

Edit Subpopulation Distribution modal showing Current Distribution and Future Distribution columns for adult and pediatric segments

Edit Subpopulation — current and future distribution

Edit Diagnosis Rate modal with adult and pediatric diagnosis rate fields and growth rate column

Edit Diagnosis Rate — adult and pediatric rates

Edit Overall Treatment Rate

Edit Payer Mix modal with Commercial, Medicare, Medicaid, Cash rows and Current and Future percentage columns

Edit Payer Mix — current and future channel split

Edit Site of Care Mix

Forecast Engine

Three levels of configuration

The engine selector exposes three method tiers. The progression was deliberate: keep the MVP approachable for the initial expert user group while giving the roadmap a clear path toward more advanced modeling.

Standard — deterministic calculation with user-provided values. Familiar to anyone who built a forecast in Excel
APEX Algorithm — a proprietary algorithm pack for more sophisticated scenario modeling
Custom Configuration — granular, line-item method selection per forecast input

Standard — deterministic, direct calculation

Custom Configuration — per-line-item method selection

Overlay tools

Four modular overlays extend the builder without leaving the main view. Each opens over the active scenario, uses its data as input, and closes back to the same builder state.

Modal patternOpens over the active scenario, closes without losing state

Portfolio Analysis modal showing Portfolio Contracting, Price Optimizer, and OpEx Analysis sections

Portfolio Analysis — contracting, price optimizer, OpEx

NPV Calculator

Run Sensitivity Analysis modal with patient population and product input variance controls

Sensitivity Analysis — variance ranges per input

Governance

Reproducibility built in, not bolted on

Scenarios in APEX are versioned artifacts. A methodology change doesn't erase a prior snapshot.

Each scenario references the baseline it was built from, the engine version, and the assumption set at creation
Sandbox-first collaboration — work privately, publish when ready, control what others can modify
Audit trail tracks who changed what, and when

Governance wasn't a reporting layer added later. It was the precondition for the tool being trusted in a portfolio planning cycle.

WhatIdesigned

End-to-end pharma forecasting prototype from RFP brief to executive pitch — no devs, fast iteration with partners, directors, and PM
Scenario library as the central organizing structure — each forecast a named, versioned artifact
Six market input modals with current/future structure and source-linked assumption fields
Three-tier forecast engine: Standard, APEX Algorithm, and Custom Configuration
Market Map covering prevalence, subpopulation, payer mix, and competitive treatment rates
Four modular overlay tools: Portfolio Analysis, NPV, Sensitivity, and Export
Governance system with versioned snapshots, sandbox vs shared workspace, and publishing controls
Rebranded and refined for executive stakeholder review as APEX

Reflection

Designing for experts who know exactly what they need

Forecasting analysts and pharma strategy leads have strong opinions about their tools. Any interface that adds steps, hides assumptions, or obscures the logic gets rejected fast.

That constraint pushed the design in one direction: transparency. Show the assumptions. Make them editable. Version everything. The goal wasn't to simplify pharmaceutical forecasting. It was to make the workflow fast enough to run during a strategy session, auditable enough to survive a methodology review, and structured enough to grow beyond Atopic Dermatitis into the broader J&J portfolio.