Google Analytics’ New Social Reporting: Do All The Dots Connect?

Google Analytics’ New Social Reporting: Do All The Dots Connect?

Yesterday, Google Analytics began releasing a new group of Social reports for current eCommerce merchants. The new reports, which will be available in the ‘traffic sources’ area, are being offered to retailers as a way to track and report return on investment for social media.

According to Phil Mui, group product manager at Google Analytics, “Marketers will be able to tie social media spend to the bottom line of the business and see how much revenue or real engagement the social channels have brought to them.”

The need to attribute conversions and revenue to specific marketing moves is obviously necessary in today’s multi-channel sales referral universe, but are Google Analytics’ new tools going far enough and providing the best information possible?

A quick look into how they’re attributing sales to social media activity leaves a few questions unanswered, especially in regards to the “Assisted Social Conversions” reporting.

Two of the data points that can be pulled from their new features include ‘Last Interaction Social Conversions’ and ‘Assisted Social Conversions.’ The Last Interaction Social Conversions report indicates when a visit from a social referral results in conversion immediately. No problem here. It’s pretty cut and dried.

What happens if a referral from a social source doesn’t convert right away, but ends up returning to your site sometime in the future and does complete the purchase? Well, this interaction is dropped into the Assisted Social Conversions pile.

Let’s think about this for a moment. Someone clicks a video link on your company’s Facebook wall, watches the video and then leaves your site. Then one month later they return and make a purchase. With Google Analytics this purchase will be tallied up in the Assisted Social Conversions column.

There seems to be a large amount of really valuable information missing here. What was the actual tipping point for that conversion? Was that person exposed to other advertisements or branded content in that month between when they visited your site originally and actually completed the sale?

To be fair this particular data point is only being touted as an “assist, and Mui says other factors may have influenced the consumer’s decision to return and buy, but that Google Analytics will record it as a “socially influenced” sale. “This way the report defines the upper and lower boundary of what we can consider socially influenced, even though it could be 10 sessions ago and happened two months ago,” he says.

But there definitely still appears to be a bit of stretching when it comes to connecting dots of this nature. In the example described above, what exactly transpired in the month between that person’s exposure to your social media and the time at which they decided to spend money with your company?

Were they exposed to a display ad while browsing the web, or maybe they read a product review on an independent blog? Did they see a commercial on TV, or run across a cleverly worded Tweet that convinced them to click through to sales content? What if the original Facebook wall video content that Google is handing the “assist” to had nothing to do with the reason that they ended up converting?

Now, it’s understandable and expected that there will be a margin of error when we begin to track and report on the influence social media may or may not have on a company’s total revenue. However, the large gap in logic that pervades Google Analytics’ Assisted Social Conversions metric is hard to overlook, especially when the technology to follow the information chain and attribute assists accurately exists.

With retargeting for example, my colleagues present clients with a “recapture” report that includes information detailing:

    – what ad was served
    – when the ad was served
    – the website on which the ad was served
    – how quickly the user returned to your website after being served the ad

Very simply, conversions are not attributed to the display ads unless the information chain can be proven.

If someone leaves that chain and comes back to your site without having been observed, is it accurate to say that an ad they may have seen or clicked on contributed to a sale? Yes, it is accurate, but it’s not exactly foolproof.

The technology to tie a unique purchase to an individual user and track the content that was a direct prompt for making that purchase exists; it just isn’t being employed by Google Analytics for this newest reporting feature rollout.

by Aubrey Beck, Content

image source: