Are you getting ROI on your Social Media activity?

Body

The inspiration for this post came through a moment of serendipity while enjoying a drink around my neighbour's pool. I was introduced to a family friend who we quickly worked out happened to also be an IBM Digital Analytics customer. It's always fascinating to hear what customers have to say and in this case I posed the question, "What business challenges are you struggling to answer?" To give you some context, this person was heading up Social Media for a business in the retail sector. The challenge faced was how to show ROI from all the social media activity they had going on.

While the eyes of most other people around glazed over, I launched into an excited run-through of campaign tagging, marketing attribution models, and reporting. Allow me to delve into these a bit more now for those of you who have continued reading to this point!

The first challenge faced by this customer was how to start measuring both paid and organic social media activity within IBM Digital Analytics. They had a small team who looked after social posts and the notion of "tagging" the posts with a marketing parameter had already been discussed, however the team was under-whelmed with the prospect of adding to their heavy workload.

The two arguments I posed to this were by going that extra mile and adding the tagging they would open up the possibility of every analyst's dream, proving ROI. With data at their fingertips it would strengthen the case for deploying more resources to a part of the business that was proving it's value. Secondly, the practice of formulating the tags can be greatly simplified by using IBM's Tracking Code Generator. This allows them to quickly and accurately create the destination URLs complete with marketing parameters. Best of all, it will flag any errors before you put the links live to ensure you have great data quality and completeness.

On the question of paid and organic posting, by using a different value for the vendor level of your MMC parameter you can distinguish Facebook Paid from Facebook Organic e.g. cm_mmc=FacebookPaid-_-Sale-_-Creative1-_-Jeans or cm_mmc=FacebookOrganic-_-Sale-_-Post-_-Jeans

The beauty of this is you can then assign these both into a Social Media channel by going to Manage > Marketing > Marketing Channels. Within here you can also define whether it is Paid, Owned or Earned media, which flows through into several of the marketing attribution reports. If you don't like the look of those URLs on your posts you can use the tracking code generator to obfuscate the parameters or even use a URL shortening service on your final URLs.

Now that the tagging is done you can jump into the really interesting part, which is marketing attribution. For those of you who have not come across this term let's quickly run through an example using our buddy Joe.

In this scenario if we only looked at same session activity we are lead to believe that organic search was the channel driving the sale. In fact our Facebook ad and an email campaign also had a part to play. So how do we give credit to them?

In Digital Analytics you have a number of attribution models that allow you to attribute the sale to First click (Facebook), Last click (Organic Search), Even (applies credit evenly across all) and lastly you can also choose a custom model made up of a mix of these three. A few reasons why you would use attribution include:

• Reassign Direct Load credit to the previous touch points in the conversion process.
• Associate site activities (signups, wishlists etc) with the programs that drove them.
• Measure the true value of marketing activities.

If we were only looking at Last Click attribution our Facebook Ads would suck because they'd never get the credit they deserved. Equally if we only considered first click it would give us an unrealistic view of how those marketing dollars are working, or as Avinash Kaushik puts it, "First click attribution is akin to giving my first girlfriend 100% of the credit for me marrying my wife."

However there is a useful calculated metric that you can apply using first click and last click to asses which channels are good at opening the sales cycle and which are good at closing sales. Using a calculated metric divide your First click Sales by Last Click Sales. Channels with a large number in the Sales First/Last Ratio column are good cycle initiators. Ensure that the messages you deliver through these channels focus on introducing new products or initiating new shopping cycles. Channels with a small number in the Sales First/Last Ratio column are good cycle closers. Ensure that the messages you deliver through these channels focus on motivating shoppers to buy. For example, an offer for free shipping is most successful when you deliver it through a channel that is a good closer.

Using Even or Linear attribution as others call it is going to apply an even credit across each channel. This is not the perfect solution since it's highly unlikely that each channel had an equal influence on someone's decision to purchase however it's a better compromise than only looking at First and Last Click. Now at this stage we're starting to get a better picture of the ROI on your social media.

As you start to get a grip on attribution you can tailor your attribution within a client ID. Go to Admin > Attribution Settings as shown here to change your settings. If you would like additional attribution metrics these can be added on to your account by contacting your account executive.

To bring your analysis to the next level you can create custom attribution metrics. Before you jump into this there are many factors you will need to consider before setting these up such as how far back should you look, deciding on what weighting to apply across first, middle and last interactions, as well as what you've learned from using the out of the box models. If this is something you'd like some help with drop us a line and we'd be more than happy to help.

Thanks,

Aidan

Further reading…

-IBM Digital Analytics Team

Curious about other presentations we've done?Click here and check under Connect With The Experts: Support Technical Exchange (STE) Replays

If you haven't visited the IBM Digital Analytics blog before, click here:IBM Digital Analytics Blog, then bookmark

If you're interested in getting a demo of IBM Digital Analytics click here:About IBM Digital Analytics