How long product certification actually takes, and how to plan for it

Published 2026-09-28 · Product Cert Desk

Ask a lab how long certification takes and you will hear that it depends. That is true and unhelpful. Here is the realistic range for most consumer products: six to sixteen weeks from a finished test plan to a certificate in hand. Simple products covered by a single standard land near six. Products with multiple standards, wireless radios, or children's safety requirements land near sixteen. Plan for the long end and you will rarely be surprised.

Where the weeks go

The timeline has five phases, and only one of them is the actual testing:

What stretches the timeline

Four things cause most delays. Failed tests that require redesign and a second round. Incomplete technical files that stall the lab's review while they wait on your documentation. Booking the lab late and landing in a queue behind every brand shipping for the holidays. And component changes mid-stream: a swapped supplier, a different dye lot, or a moved factory restarts the clock on the affected tests, because the certificate covers the exact configuration that was tested.

How to compress it

Write the test plan before the first production run, while changes are still cheap and fast. Book lab capacity as soon as the plan is signed, not when samples are ready. Prepare the technical file in parallel with testing instead of treating it as a follow-up task. Freeze the bill of materials the day samples are made, and treat any change after that as a formal decision with a timeline cost attached. Pre-test internally where you can: catching a failure on your own bench costs days, while catching it at the lab costs weeks.

Planning backwards from launch

Take your first ship date and subtract sixteen weeks. That is your test-plan deadline. Add a four-week buffer for one round of retesting, because most products need at least one. If the resulting date is in the past, you have three options: simplify the product scope, pay for expedited lab slots where they are offered, or move the launch. What you cannot do is compress someone else's queue.

Certification is a scheduling problem disguised as a technical one. The brands that hit their dates are the ones that booked the lab before they needed it.

The technical file: what the lab expects

While testing runs, the lab will ask for the technical file, and an incomplete one is one of the most common sources of delay. For most consumer product certifications, the file needs to include a general description of the product and its intended use, design and manufacturing drawings with a complete bill of materials, a list of the standards applied, test reports, user manuals and labeling artwork, and a risk assessment where the standard requires one. None of this is exotic. The delay comes from assembling it under time pressure after testing has already started.

The highest-leverage move is to start the file the same week you finalize the test plan. Assign one owner, keep it in a shared folder the lab can access, and treat every design change as a file update. When the lab asks for the bill of materials in week nine, you send a link instead of starting a two-week internal scramble. Auditors and regulators can request this file years later, so the version you submit should be the version you would defend in an audit.

Retesting: budget for one round

Most products fail at least one test on the first attempt. This is normal and should be in the plan, not treated as a crisis. A single failed radiated-emissions run or a drop test that cracks a housing typically costs two to four weeks: a week to diagnose and implement the fix, a week or more to get back on the lab's schedule, and the retest itself. Products that pass everything first time exist, but planning around them is how launches slip.

The way to keep retesting cheap is to fail early and fail small. Pre-compliance testing on your own bench or at a local lab catches the obvious problems before the accredited run. Design the product with margin above the standard's limits rather than exactly at them. And when a failure happens, fix the root cause once instead of iterating at the lab's hourly rate. Every round of retesting at an accredited lab costs more than the last one did, because the queue does not care about your launch date.