Measure and Analyze: effectiveness of a video in your website

"There is never enough time, never enough." Brokeback Mountain

Lately there are clients who ask me the same ... "videos help to conversion as it seems?".

Logically depends on the video, product or service offered, the page and where to place the video ... in short, not easy to confirm or deny the practice of using a video is good, fair or poor. But I can say something very satisfied: Measure it. And you will know without asking anyone if video works as YOU want it to work ☺

Where do we start?

First, decide / the objective / s that we pursue with the video. Say is a video that promotes a particular product or brand and what we seek is to help direct selling: conversion. It may also be that our ultimate goal is to get a lead or explain a service offered. In any case, the video you must associate one or several objectives, otherwise they will not be possible to measure them.

The KPIs that we associate with the objectives point the way to collect the data we need. The most commonly used are:

Acquisition (attention paid to video)

• Number of times it begins to view the video: it tells us how well the video is integrated into a website, how it has worked the image, use the video call, etc..

Do not forget to add context to this metric:

• % Displays / total visits to the page where the video is integrated
• tendency of both metric over time
• % Displays / total video views on the website

Hook (using video)

• Average time video display

Do not forget to add context to this metric:

• % Display time / total time Video
• tendency of both metric over time
• % Vs time video display. % Time viewing other videos from the web site

A metric that I like to use as a supplement to this engagement KPI is the number of users who viewed the entire video. Knowing this information will we know if the average time display is high because most users see it in full or because most users were about half finished and some do. It seems the same, but it is not. The first indicates a very high acceptance of the video and the second, on the contrary, a vanishing point of interest to optimize ☺

What I usually do is split the video into parts and see exactly how much may be causing rejection, tired or whatever problems need our early optimization to get really match.

• Is there a section of video to be played more than once per visit?
• What is the least section view?
• What point in the video there are more dropouts?
• If the video is displayed on more than one page ... where it works best at the level of visitation?

Do not forget a metric that can contaminate our results, the loading time of the video, because if the results are negative Down KPIs may have something to do that it takes long to load the video.

Conversion (push to achieve target)

• Number of users viewing a video and convert

Do not forget to add context to this metric:

• % Conversions with video / total conversions
• If the video is displayed on more than one page ... where it works best at the level of conversions?
• Video transaction price / average total transactions
• tendency of these metrics over time

If segmented video displayed, we have a fairly reliable statistics which videos are helping directly to conversion.

Retention

• Number of videos viewed per visit

Segmented by each of the videos on the web can help us recognize the videos that work best for retaining users and find synergies between products or ways of showing that allow us to make the most of this type of communication.

If the video we took him out to other platforms in addition to our website such as YouTube, Facebook, etc. will be very interesting to see the interest generated out of context and if the other metrics are similar (loading time, time display means, number of times shared, etc..). Many times, using social media to have more data on the performance of a video can help to detect problems or viralizarlo so that part of our online strategy.

With a web analytics tool (Google Analytics from any other payment as Webtrends or Adobe) can measure any metric of the mentioned above, even set a target associated with the video so that we can then target to deepen the previous behavior and rear of users who interact with the video:

• What are these users?
• What content pre-visualized video?
• Where did they come?
• What were to get to the video?
• What did after seeing the video?
• How do they differ from other users?
• What percentage of users left the site after seeing the video?

Thus, we can get patterns that allow us to learn where best to place the video, where to advertise it, what we can generate cross-selling, etc.. In addition, this monitoring can determine if this video helps to increase sales or not, what kind of content works best and why, what length should have, etc.. An example of a dashboard for this type of information is as follows (done in Excel):

Not enough time to measure everything, I know :)