This week in Youtubular Conversations I look at how Google Analytics can be used to learn about visitors from YouTube. I wanted to talk about paid YouTube ads but failed. My research lead a story about organic traffic from YouTube and understanding its value. This episode is also a gate-way to semi-advanced Google Analytics.

Google Analytics is NOT Scary, it’s FUN

Today professional communicators are being ask to provide qualified leads, and even sell products. That means directing folks to a web site, not once but multiple times. Even on a community or not-for-profit site where education and outreach is “the product,” it’s likely that folks who come more then once, are more likely to be activated, engaged, educated and inspired.

YouTube Is A Major Component Of Attracting Visitors Who Will Visit Again and Again – A Goal For Most Brands

Communicators need to amplify their storytelling with paid media. I wanted to compare YouTube Ads and AdWord Ads; they are very different in nature.

AdWords are triggered by searches that are often questing for answers that have a tangible immediacy like “restaurants near me open now,” “where can I eat Italian near me?,”  “best laptop under $500,” or “computer store near me open after 8pm”). In-Market searches indicate folks are close to making a purchase decision, or semantically indicate a purchase decision has been made, and they are looking at where to buy it “where can I buy an Samsung Chromebook model XE513C24-K01US .”  Ad buyers track users through there entire series of queries and pay a premium to reach folks “ready to buy.”

YouTube is considered the world’s second largest search engine, regular visitors to YouTube may rarely search. Regulars are often subscribed to specific channels and the user interface (UI) floats episodes from subscribed shows to the surface, along with curated selections based on user interests, and general trends.

YouTube is The World’s Second Largest Search Engine

Those who are searching on YouTube are not likely to be “ready to buy.” They search early in the  information research stage, or even after a purchase. A Google Millard Brown study found folks 64% of folks search YouTube regarding electronics both a week before and a week after purchase. Other searches may have nothing to do with purchasing like those looking for a rant, review, to hear a story, seeking new skills, or general education (5 Ways A Car Wash Will Destroy Your Car, or How can I make a fence without cement).

Paid YouTube

I’ve bought ads that run on YouTube direct from Google and through 3rd parties. But I was looking for some data I could share with you for this episode. As I noted above I failed. If you have data that can be publicly shared. I love to hear from you [email protected].

TGMT: Cheap Trinkets and $300 Learning Thermostats

The Google Merchandise Store (TGMS)

Google runs paid ads to drive some traffic to their merchandise store ($3k in 2016) and they make the data available through Google Analytics. This is a real store that in 2016 sold ~ 2.5 million dollars — about .01% of Alphabet’s* 2016 gross revenue (*Google’s parent).

It turns out that TGMS doesn’t buy paid spots on YouTube, at least done that I found in the data from 2016, or 2017. They do get a fair amount of traffic from YouTube, more than any other social network. They also don’t get a lot of sales directly from folks who directly visited from YouTube. The data as I discuss in this episode of Youtubular Conversations indicates that in 2016 YouTube was the largest source of organic traffic from all social media sites. YouTube wasn’t however the largest source of revenue as you can see below.

TGMS direct revenue from social media referrals (2016)

In 2016 TGMS had less than $700 in direct sales from YouTube. For this episode I examined data from July 20, 2016 to Dec 20, 2016; in that period of time there was less than $150 in direct revenue spread across seven transactions. But direct revenue isn’t the only story here.

Gini Dietrich, my friend, FIR Podcast Host, (and client) agrees that PR Can generate sales but don’t forget about ESO (Earned, Shared and Owned).

Time May Be Your Best Sales Person

At TGMS  about 70% of their sales come from folks who have visited more than once. I know because I’ve spend about 40 hours reviewing the data for TGMS in the last 6 months for a class I teach.

One thing I’m sure of (in general) and specifically for this store… it takes time to make sales. Which should make sense to communicators. It takes time to build relationships, trust, and mind share. It takes time to find the time to engage, research and transact.

Blue – All Visitors
Orange – YouTube Visitors
Green – All Multi-Session Visitors

As the chart indicates, YouTube visitors are less likely that the average visitor from returning to the site. But the they do return. Many folks who do transact on TGMS do so on their 2nd, 3rd, 4th or 5th visit.

Google Analytics has lots of amazing features but many of them only retain data for 90 days and they don’t allow for retention of any personally identifiable information.

Example Reports

In the episode I mention a number of the reports I looked at. I have included them here. Unless otherwise notes, all of these reports were recorded on Sunday Feb 5th, 2017 and is for TGMS using a data set from July 20, 2016 to Dec 20, 2016.

Shows all traffic vs. social traffic and then breaks down the social traffic by network.

We can see that visitors from the YouTube referral cohort gravitated towards particular SKUs. In the next report we can see the specific details for a single SKU.

Product (SKU) centric view – shows traffic by segment and dimensions for that SKU; in this case a pouch.

Attribution attempts to value each step in a conversion. This view shows two attribution models. One that weights the 1st interaction and 1 that weights the last interaction. When weighting the 1st interaction, YouTube contributes more to the sale.

Similar to the previous image but for a period of 45 days. How many days should your model look back? From this data and other studies we know traffic from YouTube is valued for a few weeks and then drops as the likelihood of conversion drops off.

We can see when looking back 90 days the influence or value of the first visit drops.

We can see that the vast majority of folks who add items to a shopping cart abandon it.

The same report as noted above but in this case looking all all users who have visited more than once “Multi Session Users” which is a pre-set segment within Google Analytics.

We can see that even once the checkout process starts, there are multiple steps (some would say way to many), and there is “friction” that causes a drop off. However,; 4.83% of those who start, finish which is more than double the number in comparison to all users. However, the data set is small we can’t have confidence in that result.

This report shows the next stage of the funnel.. after starting the check out process we can see that our YouTube cohort continues to drop out of the process. Making this process more friction-less would likely pay off. This also explains why repeat purchasers will have an easier time, and maybe more likely to purchase (if they do come back).

We can see that all the revenue directly from social network visits is very small (17k vs 2.5 Mil), and YouTube visitors where highly engaged (lots of conversions (goals like looking at product info, or time spent on the site, etc.) but only purchased a small amount.

Looking at all sessions (Orange) vs. Social Sessions (blue) we can see that while in some places it matches general traffic trends at other points it is highly divergent. We can see that is at its peak during the US Holiday selling season.

Many brands use page depth (the number of pages viewed per visit) as a KPI. This report shows a large number of YouTube referral traffic have only visited a few pages. There were ~ 1 million sessions total, and 400k YouTube sessions — roughly only 10% of all traffic looked at 1 page only, but 1/8 of YouTube traffic did.

If TGMS was using paid ads on YouTube that would have entered the cost here, and it would have been listed as CPC or another type of paid visit. You can see that ad costs from AdWords have been entered (line 5). ROAS stands for Return On Ad Spend, and you can see they had a 1% click through rate, and a 1049.36% ROAS; for every dollar spent, they received a 1049.36% of that cost in sales (they spent ~3k).

We can see that for all users (blue) and all multi-session users (green), folks aged 25-34 where the dominate demographic (other reports show that this age group buys the most in the store). In comparison the YouTube referral cohort is both younger and older. ~4-5% more folks aged 18 to 24, and a similar amount less age 25 to 34, and increases (over the average) as well in the 56 – 64, and 65+ range. If TGMS wanted to get more sales from this cohort they might attempt to attract more folks in the prime 25 – 34 age group.

Revenue for all of 2016.. we can see it start small in Jan and ramp up through the holiday buying season. We can see some of the best performing SKUs were Nest product. We know from earlier reports that the YouTube cohort didn’t buy these products which are far more expensive than other items on the site.

You Too Can Access The Demo Account

The demo account is open for everyone. If you don’t have a free Google Analytics account you will need to create one. If you normally access Google Analytics through a work emai/loginl, etc. you may wish to use a personal account if you don’t want the demo account showing up on the same dashboard you use at work. Click here for the instructions and link.

The post YouTube and Google Analytics appeared first on FIR Podcast Network.