Calculate drop-off rate from started and completed users, abandoned and initiated counts, or a target drop-off rate to find needed completions.

Drop Off Rate Calculator

Use the tab that matches the numbers you already have.

Started + completed
Abandoned + initiated
Target rate
people
people
people
started
people
%
Drop off rate
0.00%
Interpretation
Status
Result interpretation appears here.
Drop-offs
0
Completed
0
Completion rate
0.00%
Copy result
Copied

Drop Off Rate Formula

The calculator uses one of three formulas depending on the tab you select.

Started + completed mode:

Drop off rate = ((Started − Completed) ÷ Started) × 100

Abandoned + initiated mode:

Drop off rate = (Abandoned ÷ Initiated) × 100

Target rate mode:

Allowed drop-offs = Started × (Target drop off rate ÷ 100)
  • Started: number of users who entered or began the step.
  • Completed: number of users who finished the step.
  • Abandoned: number of users who left without finishing.
  • Initiated: number of actions or purchases that began.
  • Target drop off rate: the maximum drop off percentage you want to allow.
  • Allowed drop-offs: the number of users who can leave while still hitting the target.

The formula assumes each user is counted once per step and that the started count is the correct denominator. If you track sessions instead of unique users, the result reflects sessions, not people. Completed cannot exceed started, and abandoned cannot exceed initiated.

Each calculator mode handles a different starting point. Use the first tab when you have raw funnel counts at a single step. Use the second tab when your analytics report abandonment directly, such as cart abandonment numbers. Use the third tab in reverse: enter the traffic you expect and the drop off rate you want to stay under, and the calculator returns how many users can drop off and how many must complete the step to hit that goal.

Reference Tables

Use these as rough benchmarks. Actual rates vary by industry, traffic source, and step.

Funnel Stage Typical Drop Off Range
Landing page to product view40% – 60%
Product view to add to cart85% – 95%
Cart to checkout start25% – 50%
Checkout to purchase20% – 40%
Lead form view to submit60% – 85%
Signup to onboarding complete30% – 60%
Drop Off Rate Interpretation Action
Below 10%LowMonitor, no urgent change.
10% – 25%ModerateReview the step for friction.
25% – 50%HighPrioritize testing and fixes.
Above 50%SevereRebuild the step or its inputs.

Example Problems

Example 1: Checkout step. 1,000 users start checkout and 720 complete it. Drop off rate = ((1000 − 720) ÷ 1000) × 100 = 28%. That falls in the high range, so the checkout step is a candidate for testing.

Example 2: Cart abandonment. 280 carts are abandoned out of 1,000 initiated. Drop off rate = (280 ÷ 1000) × 100 = 28%. The completion rate is 72%.

Example 3: Planning to a target. You expect 1,000 signups and want to keep onboarding drop off under 20%. Allowed drop-offs = 1000 × 0.20 = 200. At least 800 users must finish onboarding to hit the target.

FAQ

Is drop off rate the same as bounce rate? No. Bounce rate measures sessions that leave a site after one page. Drop off rate measures users who do not advance from one specific step to the next.

Is drop off rate the opposite of conversion rate? For a single step, yes. Drop off rate plus completion rate equals 100%. Across a full funnel, conversion rate is the product of step-by-step completion rates.

Why does my drop off rate change by traffic source? Different sources send users with different intent. Paid social often shows higher drop off than branded search. Segment your data before drawing conclusions.

Should I count unique users or sessions? Use whichever your analytics platform tracks consistently. Mixing the two in one calculation will distort the result.

What is a good drop off rate? It depends on the step. A 20% drop off at checkout is normal. A 20% drop off when clicking a primary call to action on a landing page would also be normal. Compare against the benchmarks in the table above and your own historical data.