How to prove someone watched a video

This comes up more often than you might expect. A care home needs to show an inspector that every member of staff completed safeguarding training. A clinic wants a record that a patient was shown the risks before a procedure. A construction firm needs site inductions on file before anyone goes on site. An agency wants to be able to show a client was told about a change to scope.

In every case the requirement is the same. Not just that the video was sent, and not just that somebody clicked something, but that a named individual received it, watched it, and acknowledged it. And that the record still means something when it is looked at a year later by someone who was not there.

What surprised us, when we went looking, is that there is no obvious product for this. There is a great deal of software that does most of it. Very little that does all of it in one place, which is why most people end up stitching two or three tools together and inventing a process to hold them together.

Why a view count is not evidence

Most video tools report views. Views are not attributable. A view count of twelve does not tell you which twelve people, whether any of them watched past the first ten seconds, or whether one person opened it twelve times.

Watch-time analytics are better but still aggregate. Knowing that the average viewer got 78 percent of the way through is useful for improving the video. It is no use at all if you are asked whether a particular employee completed their fire safety training.

The gap between the two is what causes problems. A system can honestly report that a video was watched while being unable to tell you by whom. You find out this matters at the worst possible moment, which is when somebody asks.

What a defensible record actually contains

Whatever approach you take, a record that holds up needs most of the following. It is worth checking any tool against this list before you commit to it.

• Who. A named individual, not a device, a session, or an anonymous view.
• What. Which specific video, and ideally which version of it, since policies get updated.
• When. A date and time, recorded by the system rather than typed in afterwards.
• That it was completed. Played to the end, not opened and abandoned.
• An acknowledgement. Something the person actively did to say they watched and understood it.
• Supporting detail. An IP address, for instance. On its own it proves little; alongside a name and timestamp it makes the record considerably harder to dispute.
• Exportable. If you cannot get it out of the system as a file, you cannot hand it to anyone.

One thing worth being realistic about: no system can prove attention. A video can play to the end in an unwatched browser tab. What a good record demonstrates is that the material was properly delivered, that it ran in full, and that a named person put their name to having watched it. That is the same standard the entire e-signature industry operates on, and it is generally sufficient.

The five approaches people actually try

1. Email the video and ask people to reply

The default, because it costs nothing. Send a link, ask for a reply confirming it has been watched, keep the replies.

Works when: you have a handful of people, an informal requirement, and nobody is likely to challenge it. If you are a team of five, this is genuinely fine and you should not overthink it.

Falls down because: replies are voluntary and most people do not send them, so you spend more time chasing than sending. There is nothing linking the reply to actually watching anything. And your evidence is a mailbox, which is awkward to produce and easy to lose when someone leaves. The chasing is what scales badly. At twenty people it is a nuisance. At a hundred it is somebody's week.

2. A video platform with the controls locked down

Vimeo and similar platforms let you host privately, remove the scrub bar so viewers cannot skip ahead, disable playback speed, and see watch-time data. Some support interactive elements at the end of a video.

Works when: you mainly need people to genuinely watch the thing, and a rough sense of engagement is enough.

Falls down because: the analytics are not tied to a named person. You can prevent skipping and still not be able to say who watched. Bridging that gap means either issuing a separate link per person and tracking them yourself in a spreadsheet, or bolting on a second tool for the acknowledgement. This is where most people end up, and it is a strange place to be: two subscriptions, a spreadsheet, and a process that only works while the person who invented it is still around to run it.

3. An education tool such as EdPuzzle

Tools built for classrooms let you assign a video to named people, block skipping, and see how much each person watched. They are often free or cheap.

Works when: you want per-person progress and are comfortable with something designed for teaching.

Falls down because: the output is a progress report rather than an attestation, learners usually need accounts, and the framing is students and assignments rather than staff and compliance. Asking a care assistant to create a student account on a classroom platform to watch a fire safety briefing is a hard sell, and for an audit a classroom progress bar is not the document you want to be holding.

4. A full learning management system

The proper answer if training is a significant part of what you do. An LMS handles enrolment, completion tracking, quizzes, certificates, renewal reminders, and reporting. Many come with a library of ready-made compliance courses.

Works when: you need structured courses, assessments, certificates and expiry tracking, and you have someone whose job includes running it. If you already have an LMS, this is very likely already solved for you and you should use it.

Falls down because: it is a large commitment for a small requirement, and the weight lands on the wrong people. Setup runs to days rather than minutes, and every member of staff needs an account, a password, and a platform to learn before they can watch a nine-minute video. Pricing typically starts well above a hundred pounds a month. If what you actually have is one video that forty people need to watch this week, an LMS is a disproportionate answer. Completion is also often recorded as a tick box, which is weaker than a signature.

5. Build it yourself

Technically very achievable. The video player APIs expose an event when playback completes, so you can catch that and write a record. Standards like xAPI exist precisely for this, with a Learning Record Store to collect the results.

Works when: you have development capacity and unusual requirements.

Falls down because: the video player is the easy part. What takes the time is per-person links, expiring them after use, storing signatures, exporting reports, handling people who never complete it, and keeping the whole thing running afterwards. It also puts you in charge of personal data you now have to protect and retain properly. We know how long this takes because we did it.

The thing nobody mentions until they are using it

Something we did not anticipate: a fair number of the people who need this do not know where their videos are. They exist, but they are spread across a shared drive, somebody's laptop, an old email thread, and a USB stick in a drawer. Which version is current is a matter of opinion.

That turns out to matter more than it sounds, because you cannot prove somebody watched the right video if you are not certain which one is the right video. Whatever you choose, the version that puts your training material in one place, with a clear title and a record attached to each one, quietly solves a problem you may not have listed at the start.

Choosing between them

The decision usually comes down to two questions. Whose video is it, and how strong does the record need to be?

If you need ready-made courses on subjects you have no material for, buy an LMS with a content library. That is what they are good at, and nothing else will give you certified course content.

If the video is already yours, and the requirement is proof rather than teaching, an LMS is more machinery than the job needs. What you want is the shortest path from your video to a record with a name on it.

Scale matters too. Below about ten people, almost any of these works, and the honest answer is to pick whichever is least effort and move on. The arguments in this article start to bite somewhere around twenty-five, and by a hundred the difference between a system and a process is the difference between five minutes and two days.

Where TrainConfirm fits

We built TrainConfirm for that second case, after looking for something that did it and not finding one.

You upload your own video. Each person gets their own link, which works once. The player has no scrub bar and no speed control, and the confirmation form stays locked until the video reaches the end. They type their name, and that signature is stored with the date, the time and the IP address it came from. Their name and the date sit on screen as a watermark while it plays. Everything exports to a spreadsheet whenever you need it.

The part we care about most is what the recipient experiences, because that is where the other options lose people. The email arrives from your company name with your logo on it, not ours. There is no account to create, no password to set, nothing to install, and no unfamiliar platform to work out. They click a link, watch a video, and type their name. Most people are done inside a minute of finishing.

It is not the right tool if you need quizzes, certificates, a content library or renewal scheduling. It is the right tool if you have a video and you need to be able to prove who watched it.