Fibonacci Agile Estimation

What is Fibonacci Agile Estimation?

Fibonacci agile estimation sizes backlog items in story points drawn from a Fibonacci-like scale such as 1, 2, 3, 5, 8, 13. The widening gaps reflect how uncertainty grows with larger items.

Updated November 2026 · 3 min read


Fibonacci agile estimation is a way of sizing backlog items by assigning each one a number from a Fibonacci-like sequence, such as 1, 2, 3, 5, 8, 13 and 21, to represent its relative effort, complexity and uncertainty. Because the gaps between numbers grow as the numbers grow, the scale forces teams to stop arguing over small differences in large, uncertain items. Nobody can meaningfully tell a 14 from a 15, but most teams can tell an 8 from a 13.

The Fibonacci sequence starts 0, 1, 1, 2, 3, 5, 8, 13, 21, with each number the sum of the two before it. Agile teams use it, or a modified version, as the scale for story points. The rationale often given is that people judge relative differences better than absolute ones, so the scale's steps of roughly 60% match how precisely estimates can actually be made.

Common Fibonacci scales

ScaleValuesNotes
Pure Fibonacci1, 2, 3, 5, 8, 13, 21, 34Grows quickly at the top
Modified Fibonacci0, ½, 1, 2, 3, 5, 8, 13, 20, 40, 100Common on planning poker decks; rounds large values to signal uncertainty
Short scale1, 2, 3, 5, 8, 13Anything larger must be split before it enters a sprint

How to use Fibonacci estimation

  1. Pick reference stories. Choose a small, well-understood story and call it a 2 or 3. Other items are sized relative to it.
  2. Discuss the item. The product owner explains the story and acceptance criteria; the team asks questions.
  3. Estimate independently. Each person picks a number privately, usually with planning poker cards, then everyone reveals at once.
  4. Discuss differences. The highest and lowest estimators explain their reasoning. Often they have spotted risks or shortcuts others missed.
  5. Re-vote and agree. Repeat until the estimates converge, then record the agreed number.
  6. Split large items. If an item lands at 13 or higher, most teams split it before committing to it in a sprint.

Fibonacci estimation example: From points to a sprint plan

Consider a hypothetical team whose last three sprints completed 26, 30 and 28 points, an average velocity of (26 + 30 + 28) ÷ 3 = 28 points. At sprint planning, the top of the backlog holds stories estimated at 5, 3, 8, 2, 5 and 3 points, totaling 26 points, followed by an 8-point story. Adding the 8 would bring the total to 34, well above the team's average, so the team commits to the first six stories and keeps the 8-pointer as a stretch item.

A seventh story further down was estimated at 21 points. Rather than carry it as is, the team splits it into stories of 8, 5 and 5 points. The new total of 18 is lower than 21. The sum of split stories often differs from the original estimate, in either direction, because the team now understands the work better; the team re-plans from the new numbers rather than forcing them to add up to 21.

Fibonacci vs. other estimation scales

  • Linear scale (1–10). Implies equal precision across sizes, which invites false precision for large items.
  • T-shirt sizes (XS–XL). Quick and intuitive for early roadmap sizing, but harder to sum into a forecast. Many teams use T-shirt sizes for epics and Fibonacci for stories.
  • Hours or days. Easy to understand outside the team but tend to be treated as commitments and vary with who does the work.

Common pitfalls

  • Converting points to hours. A fixed conversion turns relative estimates back into time estimates and loses the benefit.
  • Comparing teams by points. Each team's scale is its own; a 5 on one team means nothing to another.
  • Averaging votes instead of discussing. If votes are 3 and 13, the average hides a real disagreement about scope.
  • Re-estimating finished work. Adjusting points after the fact to match actual effort distorts velocity data.

Fibonacci estimation is one of several approaches to estimation in agile teams. The resulting points feed sprint velocity and sprint planning.

Frequently asked questions

Why do agile teams use the Fibonacci sequence for estimation?

The widening gaps between Fibonacci numbers match how uncertainty grows with larger work. People judge relative differences better than absolute ones, and nobody can meaningfully separate a 14 from a 15, but most teams can tell an 8 from a 13. The scale stops teams from debating small differences in large, uncertain items.

What is the modified Fibonacci scale used in agile estimation?

The modified Fibonacci scale runs 0, ½, 1, 2, 3, 5, 8, 13, 20, 40 and 100. It keeps the Fibonacci pattern for small values but rounds large ones, which signals that big estimates are rough. It is common on planning poker decks, while some teams use a short scale ending at 13 and split anything larger.

Should Fibonacci story points be converted into hours?

No. A fixed conversion turns relative estimates back into time estimates and loses the benefit of Fibonacci sizing. Points describe effort, complexity and uncertainty relative to a reference story, and each team's scale is its own. Use points with sprint velocity to forecast how much work fits in a sprint instead.