TapCub is live — analytics, insights and live chat, free on one platform
Glossary

Annual Growth Rate: Definition, Formulas and a Worked Example

Annual growth rate is a family of measures — YoY, annualized and CAGR — that give different answers on the same data. Formulas side by side, one dataset worked four ways, and how to annualize monthly growth.

Su He Reviewed by Oliver Reed Also called AAGR Last updated Nov 16, 2026 Glossary

Quick answer

Annual growth rate is the percentage change in a metric over one year, but the name covers at least three formulas: year-over-year change, an annualized short-period rate, and the compound annual growth rate (CAGR). Always state which formula produced the number.

Formula: CAGR = (end ÷ start)^(1 ÷ years) − 1

What is annual growth rate?

In reporting, annual growth rate is the percentage change in a metric over one year — but the phrase covers a family of measures, not a single number. Depending on the formula, "annual growth rate" can mean the year-over-year (YoY) change between two annual values, a short-period rate annualized to a full year, or the compound annual growth rate (CAGR) across several years. On the same data these formulas give different answers, so a reported number is ambiguous until you name the formula behind it.

That matters in practice. One team quotes "+40% annual growth" from a single strong quarter, another reports the three-year CAGR, a third averages the yearly changes — all describing the same business with different numbers. Below, the formulas sit side by side, one dataset runs through all of them, and each measure's failure mode gets a number. For the single-year comparison on its own, see year-over-year growth; for month-level compounding, see compound monthly growth rate.

A quick decision rule: a one-year window calls for YoY; projecting a quarter's pace into a full year calls for an annualized rate; summarizing a multi-year path calls for CAGR.

The formulas side by side

Four measures travel under the "annual growth" label. The table gives each formula, the question it answers, and its main trap.

MeasureFormulaQuestion it answersUse it forMain trap
Year-over-year (YoY)(this year − last year) ÷ last yearHow much did this year change vs last year?Single-year performance; removes seasonalityOne year only; an unusual base year exaggerates or hides change
Annualized rate(1 + r_period)^m − 1; official form [(end ÷ start)^(m/n) − 1] × 100 (BEA)What would a full year look like at this short-period pace?Quarterly/monthly flash reports; official GDP annual ratesAssumes the pace continues; misleading for volatile series
CAGR(end ÷ start)^(1/n) − 1What constant yearly rate compounds start into end over n years?Multi-year trends in users, revenue, investmentsSmooths away the middle; sensitive to start and end points
AAGR (arithmetic)(r₁ + r₂ + … + rₙ) ÷ nWhat is the simple average of the yearly rates?A quick multi-year averageIgnores compounding; overstates growth when rates swing

Here r_period is the growth rate of one period, m is the number of periods per year (4 for quarterly data, 12 for monthly), and n is the number of periods (or years) between the two values. The annualized row is the official BEA convention: its FAQ on average annual growth defines the rate as a variant of the compound interest formula, r = [(GDPt ÷ GDP0)^(m/n) − 1] × 100, and the NIPA handbook, Ch. 4 appendix prints the same formula with a worked example. The AAGR row follows the finance convention used by Wall Street Prep: the arithmetic mean of period rates, with no compounding.

Two naming warnings before the example. First, "annualized" and "compound annual" are not synonyms: annualizing projects a short pace forward, while CAGR summarizes a completed multi-year path. Second, the name AAGR itself is ambiguous — the mistakes section below shows why. And annualizing a growth rate is not the same as annualizing a level: one month's revenue × 12 is an annual run rate, a size estimate rather than a growth rate.

Worked example: one dataset, four answers

Sample data: a fictional product's year-end active users, chosen so every step can be recomputed by hand.

Year-endActive usersYoY change
202310,000—
202414,000(14,000 − 10,000) ÷ 10,000 = +40.0%
202512,600(12,600 − 14,000) ÷ 14,000 = −10.0%
202617,640(17,640 − 12,600) ÷ 12,600 = +40.0%

Now the four "annual growth rates" for 2023→2026:

  • YoY (2026): +40.0%. Answers "how did the last year go" and says nothing about the dip in between.
  • CAGR: (17,640 ÷ 10,000)^(1/3) − 1 = 1.764^(1/3) − 1 ≈ 1.208276 − 1 = 20.83% — the constant rate that turns 10,000 into 17,640 in three years.
  • AAGR: (40% − 10% + 40%) ÷ 3 = 70% ÷ 3 = 23.33% — the arithmetic mean of the three yearly rates.
  • Total growth ÷ years (wrong): (17,640 ÷ 10,000 − 1) ÷ 3 = 76.4% ÷ 3 = 25.47% — ignores compounding and matches none of the measures above.

Same four numbers, four different stories: 40%, 20.83%, 23.33%, 25.47%. The 2.5-point gap between CAGR and AAGR is the price of the −10% year: arithmetic averaging rewards the volatility that compounding punishes. If a dashboard shows "annual growth rate: 23.33%" without naming AAGR, a reader who assumes CAGR overstates the trend by 2.5 points, and a reader who assumes YoY is off by 16.7 points.

How to annualize monthly or quarterly growth

To annualize a short-period rate, compound it over the number of periods in a year: (1 + r)^m − 1, with m = 12 for monthly rates and m = 4 for quarterly rates.

  • Monthly: 2% per month annualizes to 1.02^12 − 1 = 26.82%, not 12 × 2% = 24% — the 2.82-point gap is compounding, not rounding. Each month grows on the previous month's base — the same logic as compound monthly growth, and the reason growth curves bend upward (exponential growth).
  • Quarterly: +10% in Q1 annualizes to 1.1^4 − 1 = 46.41%.

When you have levels instead of rates, use the BEA form [(end ÷ start)^(m/n) − 1] × 100 with n = 1 for a single quarter or month. The NIPA handbook example: a component rises from 100 to 105 in one quarter (+5% at a quarterly rate), so the annual rate is (105 ÷ 100)^4 − 1 = 21.6%. BEA publishes most quarterly GDP changes this way — see its FAQ on annual rates — so a headline like "GDP grew 4.0% in Q4" states an annualized pace, not the year's total. For month-over-month mechanics, see month-over-month growth.

One caution travels with every annualized number: it is a projection, not a result. The next section shows how far off it can be.

Common mistakes

  • Annualizing a partial year. Continue the example: Q1 2027 users rise from 17,640 to 19,404 (+10%), which annualizes to 46.41%. If the next three quarters are flat, the year ends at 19,404 and actual YoY growth is (19,404 − 17,640) ÷ 17,640 = +10.0%, not 46.41% — a 36.4-point gap that rests entirely on the flat-quarters assumption. BEA warns that annual rates "may be misleading" for volatile series, and it deliberately does not annualize monthly series such as personal income because doing so "excessively exaggerates" month-to-month changes.
  • Calling AAGR a compound rate. In the example AAGR is 23.33% while CAGR is 20.83%; the two coincide only when every year grows at the same rate. The name itself is a trap: Eurostat's official glossary defines AAGR as the compound annual rate (what finance calls CAGR), while finance educators define AAGR as the arithmetic mean. Check which convention a source uses before comparing numbers.
  • Dividing total growth by years. 76.4% ÷ 3 = 25.47% looks like an annual rate but ignores compounding; the correct compound answer is 20.83%.
  • Ignoring the base. YoY from an unusual base (an outage year, a one-off campaign) exaggerates or hides change; CAGR shifts when you move the start or end year even if the middle is unchanged.
  • Comparing two official numbers for the same year. The St. Louis Fed's two-measures analysis shows U.S. real GDP grew 2.5% Q4/Q4 but 2.8% year-over-year in 2024 — both correct, different windows. Q4/Q4 compares end points; Y/Y compares annual averages. Label the window when you report.

How to measure annual growth in TapCub

Run a series like 10,000 → 14,000 → 12,600 → 17,640 through TapCub and the rule of this page still holds: TapCub gives you the levels and one built-in comparison, and you name the formula behind every rate you quote.

  1. Pull the yearly series. Web analytics › Reports lays the daily aggregates out as one row per time bucket, and the bucket can be month or year (yearly needs a date range of at least a full year). Pick columns such as sessions, conversions or revenue and export the table. One caveat: a bucket's visitor total sums daily visitors, so one person active on three days counts three times. Decide whether you are growing visits or people before you quote a rate.
  2. Read YoY from the header comparison. Set the date range to a full year and the comparison to previous period. The previous period is the year right before, so the change shown next to each number is a YoY answer of the (17,640 − 12,600) ÷ 12,600 = +40% kind. The header offers only off, previous period or same period last week; there is no "same month last year" option, so for December vs December pick each month as a custom range and compare the two numbers yourself (Compare date ranges).
  3. Compute CAGR and annualized rates yourself. TapCub has no CAGR or annualize button. Put the exported yearly values into (end ÷ start)^(1/n) − 1, a quarter's rate into (1 + r)^4 − 1, and write the formula name next to each result.
  4. Annotate the spikes. Both +40% years in the example coincide with campaigns. Add them in Configuration › Annotations with the type Campaign or Release; they appear as dashed lines or shaded bands on Web analytics › Overview and Web analytics › Reports, so nobody annualizes one campaign quarter into a 46.41% "trend".
  5. Track revenue against a yearly target. Revenue comes from purchases on the order ledger in All-platform analytics › Business › Commerce & revenue (see revenue growth rate). All-platform analytics › Business › AARRR growth shows year-to-date progress against an annual target you set for new users or revenue. Bots are counted separately in Web analytics › Bots, so crawlers do not inflate the base of any of these rates.

Sources

Was this explanation clear?

We use your answers to fix and expand this page.

Su He

Data Analyst

Explores traffic sources, conversion funnels, and retention, using clear metric definitions and concrete examples to explain what the numbers mean.

FAQ

Questions people ask

Short answers to the follow-up questions that come up most.

Still have a question?

Chat with the team behind TapCub — we usually reply within a few hours on business days.

What is the difference between YoY growth and CAGR?

YoY compares one year with the previous year: (this year − last year) ÷ last year. CAGR is the constant yearly rate that compounds a start value into an end value over several years: (end ÷ start)^(1/n) − 1. On our sample data the 2026 YoY is +40% while the three-year CAGR is 20.83% — same series, different questions.

How do you annualize a monthly growth rate?

Compound it over 12 months: (1 + monthly rate)^12 − 1. A 2% monthly rate annualizes to 1.02^12 − 1 = 26.82%, not 12 × 2% = 24%, because each month grows on the previous month's base. For quarterly rates use the same formula with 4 periods: +10% in a quarter becomes 1.1^4 − 1 = 46.41%.

What is a good annual growth rate?

There is no universal benchmark: it depends on the metric, the base size and the window. Compare like with like — same formula, same period, same metric definition — against your own history or a named peer. And name the formula: a 23.33% AAGR and a 20.83% CAGR can describe the same three years.

What is the difference between AAGR and CAGR?

AAGR is the arithmetic mean of yearly rates and ignores compounding; CAGR is the geometric, compound rate. With volatile years AAGR runs higher: our example gives 23.33% AAGR vs 20.83% CAGR. Note that Eurostat uses AAGR to mean the compound rate, so check the source's convention.

See this number on your own site

Add one script tag for web and apps, and the first reports fill in within minutes.

See your data clearly. Find your growth.

Every click, backed by data. Install one line of code and see your first numbers in a minute.

No credit card · Free plans for analytics and chat · Cookieless analytics