Return on investment (ROI) formulas, the value of an organic visit, a like, or a retweet – over the last few years I have seen a barrage of metrics and calculations to signify the success of a campaign. From the traditional Marketing Mix Modeling (MMM) reads to simple e-commerce “per-session value,” they all have merit but they also share the same flaw. In most cases, they stop measuring after the purchase or action is completed. But does the transaction and its measurement, or consumer interaction with the brand, really stop there?
Over the past few months we have began to truly merge our search and social practices, which has led to some very interesting discussions and brainstorming sessions. Search people are trying to solve social media challenges with search solutions and vice versa; this has created a very interesting framework for innovation. One of the results of this is what I call Social Post Purchase Analytics*.
It’s a really simple concept I would love to see integrated into today’s social, analytics, and bid management platforms so we don’t have to manually “duct tape” it all together anymore. (Hint to Kenshoo, Marin, AdWords, Omniture, etc.) All these areas would come together cohesively within one campaign instead. It makes social posts part of the sales process – after the original transaction is done.
Bringing Social and Search Together to Better Understand Sales Lift
Traditionally in e-commerce, you would look at your e-commerce report and see something like this:
This is great; it tells us on a very detailed level, for each campaign/medium, what the average per-session value was, therefore allowing us to calculate how we are doing from an ROI perspective and what a bidding strategy should look like.
As an example, the data tells us that the “per-session value” for our Facebook ads is $1.23 and therefore much lower than it is for paid search ($1.63) or organic search ($3.26). Most people would just leave it at this and do a calculation against the profit and CPC/CPM. Fortunately, we found out that this is actually not quite where it ends. And let me tell you why.
The checkout flow on this (and most other e-commerce sites) looks similar to this:
Based on our ideas, we implemented two unique aspects across some of our clients’ sites; the first one is to make their social profile URLs part of their customer profile during checkout, and the second one is adding a Share widget on the Thank You page. The Share buttons inside this widget have a unique URL that is tied to this specific shopping session by retaining the original campaign parameters. The Share widget looks very similar to Amazon’s:
With these two components in place, we are now able to measure the number of shares we get after each purchase. By tying each social share to a unique checkout event, this immediately gives us a great new value I call “sociality*,” which simply measures how social (likely to share) someone is. When we tie this to an analytics event, we can create nice comparisons against it.
We already know from our events report above that people from Facebook (11.12 percent) are more likely to share a purchase than people that came via search (6.47 percent). Now we have an additional value parameter (and a second key performance indicator [KPI]) that we can use to calculate an appropriate PPC/CPM bid based on potential viral lift.
But we can even take this one step further, by then tagging the shared links with the originating campaign information; we are now able to track and evaluate purchases based on that URL and see the effects of a buyer’s sociality.
Now we can create very interesting models that looks something like this:
Out of 1,000 Facebook visits from campaign ID123, we can expect on average 201 direct sales, 88 indirect sales through sharing, and an additional 32 sales through secondary sharing. So even if the cost-per-click (CPC) for Facebook is higher than it is for paid search, in the end we are getting a larger number of sales through Facebook, justifying a higher CPC.
With that methodology in mind, you can develop much better bidding models by better understanding the true value of a consumer and the potential after-purchase leads this consumer might generate. And if you look closely at your data, you might find that the value of a visitor from social is only low on the surface and here’s why: We could monitor the supplied social media profile of the consumer to see if the consumer is discussing his purchase after the sale and evaluate the reach of that conversation (how many people saw and or engaged with it).
I know this methodology still needs to be tested and documented further, but in its early stages we have seen lots of success by simply tagging social actions and buttons, and gathering additional insights that can help us make smarter, more efficient media buys by having better ROI calculations through deeper data. Tying social shares with specific online purchasing events, and then measuring the social post-sharing analytics, offers up a perfect A/B testing scenario for marketers as well.
We have found some verticals that actually had larger secondary sales than primary ones thanks to social sharing. In these cases, someone purchases a product, shares the purchase on social media, and then multiple social connections have made purchases triggered by the initial purchase. If I were naming things, I would call this “e-commerce virality*.”
I would love to hear your thoughts on this approach. Is it something you’d consider implementing? If you are already doing it, what are your findings?
* I was asked by our marketing department to never again name things, so the official name is pending.
Homepage image via Shutterstock.
Click-through rates for a business website fall with its position in organic search results. But what is the effect when organic results are pushed further and further off screen by paid ads, Google My Business listings and Knowledge Graph?
When you’re just starting out as a business owner it’s easy to become wrapped up in the seemingly endless number of metrics ... read more
Visual search on the web has been around for some time. In 2008, TinEye became the first image search engine to use ... read more