The new Cost Data Import capability for Google Analytics offers the ability to upload not just ad costs, but also clicks and impression counts. While much of the obvious benefit comes from uploading cost information from Bing adCenter, Facebook Ads or LinkedIn Ads, you don’t have to stop there. As an example, this article explores an integration of YouTube video views into Google Analytics to demonstrate the effectiveness of that channel.

If you think of your videos on YouTube as free advertisements, then viewing a video on YouTube would be an ad ‘impression’, and you can start to see how a simple import of video views per day will provide a quick measure of YouTube effectiveness as a sales channel. We will upload these impressions to the same medium/source as our YouTube referral visits (medium=referral,, so they can be reported together.

The same could be done with Facebook page views, or any other social media source where you can get a count of ‘eyeballs’ that saw your content. You just need to make sure that Google Analytics records visits as referrals from that source, and that your upload matches the source. Note that this technique does not involve any special tagging, nor will it provide video-level (ad/campaign) effectiveness measures, but it can be done at any time and works with your historical data.

Using NEXT Analytics to download video views

You can reference our User Guide to see the steps needed to prepare data for uploading into Google Analytics. With the YouTube connectivity in NEXT, we can easily download the video views over the past few days, weeks or months. To get going, just pull down a week or so. YouTube supplies a lot of extra fields and detail that we don’t need in this case. We can do a big historical upload once things are fully configured, and there is no sense suffering long delays waiting for the data while you are setting things up.

Then select the columns you want to retain (Exclude the rest) and rename them as required. In this case, I mapped the Region to ga:adGroup, Title to ga:campaign, Views to ga:impressions, and I made sure the Date was imported as text in the correct format (2012-11-29).

I saved the results to a CSV file by adding the command through the Analyze tab, and then saved the query.

For the second part of the exercise, I create a GA Upload workbook to load those results into Google Analytics, using the values of medium (referral) and source( to match the referral visits already in GA. When the upload is complete, the resulting script commands are saved to the workbook.

To pull the two parts of the project together, I copy the script commands from the GA Upload workbook into the same workbook I initially used to download the YouTube results. Now I can adjust the value for the [PERIOD] to the date range of my historical period of interest, save and refresh the workbook. That will download all the YouTube visits and upload them as impressions into Google Analytics [and it could take a very loooooong time].

As a final step, I can set the [PERIOD] to ‘Past 3 days’ and run this daily (it allows for weekends or the odd forgotten day, and the update is set to replace previous uploads so there is no worry of double-loading).

Now we can make a simple report from Google Analytics to query the referrals from, and display the values for impressions, visits, goal conversions and even transaction revenue. Pick a date range and display the results as a chart. Of course, you could also go a step further with NEXT and look at conversion rates, or period-over-period change – all the things the NEXT does well.

Don’t stop at the numbers, though. Remember to put the results into context with the proper annotations and observations about what happened during the period that could have contributed to the result you are showing. Since NEXT automated the analysis, you have lots of time to add value.