Law 10/2025 three-minute indicator analyser

How to use it

Export the detailed call log of your customer service queue from your phone system, for anything from one day to one month, and keep three columns with this header:

• entry: date and time the call reaches your number, as 2026-11-03 10:02:15 or 03/11/2026 10:02:15. This is when it reaches the number, not when it joins the queue.
• wait_s: seconds from entry until a person answers, or an assistant that genuinely offers to speak to a person. If the call was not answered, the seconds until it ended.
• outcome: answered, abandoned (the caller hangs up) or system (the phone system ends it, for example when the maximum wait runs out or the queue overflows).

Paste the CSV into the box or load the file. Semicolons, commas and tabs all work as separators, and the Spanish header is also recognised (entrada, espera_s, resultado). If your system gives the answer time instead of the wait, calculate the column in a spreadsheet by subtracting the entry time from the answer time. The “Load a sample log” button shows the format and the result with made-up data.

What it calculates and which rule it uses

Article 10.3 of Spanish Law 10/2025 requires the companies within its scope to answer 95% of calls, on average, in less than three minutes. The law does not set the averaging period or how abandoned calls are treated, so the tool applies the demanding reading explained in the guide to Law 10/2025 on the phone (in Spanish), the same one used in the Excel self-assessment template (in Spanish):

Indicator = calls answered by a person in under 180 s from entry ÷ (answered + abandoned after the cut-off + ended by the system)

• The clock starts at entry: the menu and recorded messages use up part of the three minutes.
• Abandoned calls count in the denominator, except those that hang up before the cut-off you set, which are usually misdials.
• Calls ended by the phone system always count as unanswered. The tool also flags them: article 8 of the law forbids hanging up on customers when the wait runs out.
• It calculates the indicator for the period, for each day and for each one-hour slot, and lists the five worst slots with at least five calls.

The 95th percentile wait is calculated over answered calls: only one in twenty answered calls waits longer. It does not replace the indicator, because it leaves out abandoned calls.

Worked example

The sample log is made up: 553 calls over three working days, between 9:00 and 18:00, with more load between 10:00 and 13:00 and a 35 to 45 second menu. On Monday 2 November the team is short of agents between 10:00 and 12:00.

Over the period there are 501 calls answered in under 180 s, 521 answered, 26 abandoned after the 10 s cut-off and 3 ended by the system: 501 ÷ (521 + 26 + 3) = 91.1%. 3 calls abandoned before 10 s are excluded. The worst slot is Monday from 11:00 to 12:00, at 39.3% of 28 calls: that is where the staff shortage and the three calls ended by the system are concentrated.

Wednesday stays at 94.6% without any serious problem. It is one call in a hundred short of answering in time. That is the real margin of many teams that staff for “roughly 95%”.

How to read the result and its limits

The result depends on the log measuring entry and answer correctly. Many phone systems store the time the call joins the queue rather than the time it reaches the number. If yours does, add the menu duration to each wait, or the indicator will look better than it is. Article 22 requires the audit to check that the measurement error does not exceed 5%, so keep the original log you calculate from.

The tool does not decide whether your company is covered by the law and is not legal advice. Nor does it look at what the automated assistant said: if a bot picks up after two seconds but does not let the caller ask for a person, that call is not “answered” in the sense of the law, and you should log the wait until the person. To staff the slots that fail, use the Erlang C calculator with the “Law 10/2025” target; to understand the denominator, see the guide to calculating telephone service level (in Spanish).

Common mistakes

• Measuring from when the call joins the queue. The menu counts within the three minutes.
• Excluding long abandoned calls or changing the cut-off from one month to the next. It improves the percentage without improving the wait. The guide to the queue abandonment rate (in Spanish) explains how to justify a cut-off.
• Looking only at the monthly average. A month that meets the target with a Monday at 83% does not protect you from a complaint by someone who called that Monday.
• Keeping a maximum wait that hangs up the call. Review the queue settings and their wait limit (in Spanish) and offer a callback instead.

The values you enter are calculated in your browser and are not sent to CallsIQ. Google Analytics records visits and that the tool was used, never the values, as described in the privacy policy.