Initiative vs Epic vs Story vs Task: A Clear Hierarchy for Your Roadmap

Few topics generate as much confusion in product teams as the hierarchy of work items. Initiative vs epic. Epic vs story. Theme vs initiative. Task vs sub-task.

Ask ten product managers to define these terms and you’ll get twelve different answers. Jira doesn’t help—its flexibility means every organisation implements these concepts differently.

Here’s my honest take: the specific hierarchy matters less than having ONE consistent hierarchy that everyone agrees on.

The Common Hierarchy

While definitions vary, here’s the hierarchy most teams converge on:

Theme / Strategic Goal
    └── Initiative / Objective
        └── Epic
            └── Story
                └── Task / Sub-task

Let’s define each level:

Theme or Strategic Goal

The highest level. Business or strategic outcomes that might span a year or more.

Examples:

• “Become the market leader in SMB segment”
• “Achieve 95% customer retention”
• “Expand into European markets”

Themes live at the company strategy level. They’re too high to be actionable directly but provide context for everything below.

Initiative or Objective

A measurable outcome that contributes to a theme. Typically 1-4 quarters in scope.

Examples:

• “Increase trial-to-paid conversion from 15% to 25%”
• “Reduce customer support volume by 40%”
• “Launch localised product in Germany”

This is the level that belongs on a product roadmap. In RoadmapOne , these are your Objectives.

Epic

A significant body of work that contributes to an initiative. Usually 2-8 weeks of effort.

Examples:

• “Redesign onboarding flow”
• “Build AI-powered chatbot”
• “Implement German payment providers”

Epics bridge strategy and execution. They’re specific enough to plan around but broad enough to leave solution space.

Story

An implementable piece of work that a developer can complete in a sprint. Usually 1-5 days.

Examples:

• “As a user, I can see my progress through onboarding”
• “Chatbot responds to top 10 FAQs”
• “Accept SEPA Direct Debit payments”

Stories are the main unit of work for delivery teams.

Task or Sub-task

The smallest unit. Technical work needed to complete a story. Usually hours, not days.

Examples:

• “Create progress bar component”
• “Write FAQ response templates”
• “Configure SEPA payment webhook”

Tasks are for team coordination, not roadmap planning.

Connect your work hierarchy to strategic outcomes—sign up free

Why the Confusion Exists

Reason 1: Jira’s Flexibility

Jira lets you create whatever hierarchy you want. Epics containing epics. Stories containing stories. Custom issue types at every level.

This flexibility is powerful but creates inconsistency. Your Jira instance probably has archaeology layers from different eras of process.

Reason 2: Different Frameworks, Different Terms

SAFe uses “capability” and “feature.” Scrum talks about “product backlog items.” LeSS has “requirement areas.” Each framework brings its own vocabulary.

Teams adopt bits from multiple frameworks and end up with hybrid terminology that nobody fully understands.

Reason 3: Scale Varies

What counts as an “epic” at a 5-person startup is a “task” at a large enterprise. The labels don’t scale consistently.

Reason 4: The Middle Layers Get Squeezed

Themes and stories are usually clear. But the middle layers—initiative, epic—often collapse together or expand unpredictably.

What Actually Matters

Here’s what actually matters for effective product planning:

1. Roadmap Items Are Outcomes

Whatever you call them—initiatives, objectives, OKRs—the items on your roadmap should be measurable outcomes, not features or deliverables.

“Improve retention” ✓ “Build recommendation engine” ✗

For more on this, see outcome-based roadmaps .

2. There’s a Bridge Between Strategy and Execution

You need something between “company goal” and “developer task.” Call it epic, call it initiative, call it Fred—just have it.

This middle layer is where strategy becomes concrete enough to plan but abstract enough to leave solution space.

3. Everyone Uses the Same Definitions

The specific definitions matter less than consistency. If your team has decided that “epic” means “2-6 week body of work”—stick to it. Don’t let one PM use “epic” for a 2-day item while another uses it for a 6-month initiative.

4. The Levels Connect

Each level should clearly connect to the level above and below:

• This story contributes to this epic
• This epic serves this initiative
• This initiative advances this theme

If you can’t trace a story up to a strategic theme, something is disconnected.

RoadmapOne’s Approach

In RoadmapOne’s model , we use:

The roadmap contains Objectives (measurable outcomes) and Key Results (indicators of progress). These map to work in Jira, where teams manage the tactical breakdown.

This keeps the roadmap at the right altitude—strategic enough for leadership, specific enough for planning—while Jira handles execution detail.

A Practical Example

Theme: Grow revenue by 40%

Initiative/Objective: Increase average order value from £45 to £60

Epics:

• Cross-sell recommendations on product pages
• Bundle discount features
• Premium product tier launch

Stories (for first epic):

• Display “customers also bought” on product page
• Build recommendation algorithm
• A/B test recommendation placement
• Track recommendation click-through

Tasks (for first story):

• Design recommendation card component
• Create API endpoint for recommendations
• Build frontend component
• Write unit tests

The roadmap shows the Objective (“Increase AOV to £60”) and which team is pursuing it.

Jira shows the epics, stories, and tasks that comprise the current approach.

Common Mistakes

Mistake 1: Too Many Levels

Some teams create 7-8 levels: theme, objective, initiative, capability, feature, epic, story, task, sub-task.

Nobody can keep track. Levels blur together. Planning becomes bureaucratic.

Fix: 4-5 levels maximum. Theme → Initiative/Objective → Epic → Story → Task.

Mistake 2: Skipping the Middle

Some teams have only “big strategic goals” and “stories.” There’s no middle layer to bridge them.

This creates a gap between what leadership wants and what teams are building—the same disconnect we see when teams confuse roadmaps with backlogs .

Fix: Always have that middle layer (epics/initiatives) that makes strategy concrete.

Mistake 3: Features Masquerading as Objectives

“Launch new checkout” is not an objective—it’s a feature. “Increase conversion to 25%” is an objective.

When features sit at the initiative level, you’ve lost outcome focus.

Fix: Check each initiative with: “How will we know if this succeeded?” If the answer is “it shipped,” it’s not an objective.

Mistake 4: Inconsistent Sizing

One team’s epic is another team’s story. When teams collaborate, the mismatch creates confusion.

Fix: Define size guidance. “An epic is 2-8 weeks of work for one team.” “A story is completable in one sprint.”

Align your work hierarchy with RoadmapOne—free sign-up

The OKR Connection

If you use OKRs, they map naturally to this hierarchy:

Your OKR Objective IS your roadmap initiative. Key Results help you track whether the work is achieving the objective.

For more on connecting OKRs to roadmaps, see OKR vs Roadmap .

What To Do If Your Hierarchy Is a Mess

Step 1: Document What You Have

Before changing anything, document your current hierarchy. What does each term mean in practice at your company?

Step 2: Identify the Pain Points

Where does confusion happen? Where do levels collapse or expand inconsistently? Where are people using the same term differently?

Step 3: Define a Target State

Pick a simple hierarchy (4-5 levels). Define each level clearly. Include sizing guidance.

Step 4: Migrate Incrementally

Don’t try to reclassify everything overnight. Apply new definitions to new work. Let old work age out.

Step 5: Enforce Through Review

In planning and grooming, ask: “Is this really an epic? Or is it a story?” Gentle enforcement creates consistency over time.

Frequently Asked Questions

“What’s the difference between initiative and epic?”

The most common confusion. Here’s my rule:

• Initiative: Outcome to achieve (measurable, appears on roadmap)
• Epic: Body of work to do (deliverable, appears in backlog)

“Improve conversion” is an initiative. “Redesign checkout” is an epic that might contribute to that initiative.

“Should epics span multiple sprints?”

Usually yes. A good epic is 2-8 weeks of work. If it’s completable in one sprint, it might be a story. If it’s longer than a quarter, consider breaking it down.

“How do initiatives relate to quarters?”

An initiative might span 1-4 quarters. Within each quarter, you’ll have epics and stories that advance the initiative. The roadmap shows which quarters each initiative is active.

“What about themes? Do we need them?”

Themes are optional. In smaller organisations, company goals serve this purpose. In larger ones, themes help group related initiatives. Don’t add complexity you don’t need.

“How does this work with Jira?”

Jira’s “epic” is typically the middle layer. Jira “stories” are your execution units. You can create custom issue types for initiatives or use Jira Advanced Roadmaps for higher-level planning—but consider a dedicated roadmap tool for strategic work.

See our Jira integration guide for more.

Conclusion

The hierarchy of work items—initiative, epic, story, task—is a source of endless confusion. But the solution isn’t finding the “correct” definitions. It’s establishing consistent definitions within your organisation.

What matters:

• Roadmap items are outcomes, not features
• There’s a bridge layer between strategy and execution
• Everyone uses the same definitions
• Levels connect to each other

Get those right, and the specific labels matter much less than you think.

Stop debating terminology. Start shipping product.