Most of my work these days revolves around help­ing clients put their data together into a full 360° multi-channel view in Adobe Insight, and then use that data to drive real busi­ness results.

I thought it would be inter­est­ing to show­case some exam­ples of how mar­keters are using Insight, in Customer-centric datasets, to make real mar­ket­ing decisions.

Why would we do this? Think about how you inter­act with brands that exist online and offline.

You might use a search engine to research a new com­puter — using both paid & organic search results as you do your research, then end up at an elec­tronic retailer’s web­site and not pur­chase a com­puter. You look like a loss — you con­sumed paid search invest­ment and made cer­tain key­words appear to “under per­form.” You even explore some of the edu­ca­tional mate­r­ial that a man­u­fac­turer spon­sors on the retailer’s web­site, but you still don’t buy the com­puter.

A few days later, how­ever, after some research, you walk into the retailer’s store, and buy not just the com­puter, but a cart full of acces­sories so that you go home ready to roll.

The pur­chase path didn’t end on the web­site. The pur­chase didn’t even hap­pen on the web­site! But there was a pretty big con­ver­sion that happened.

How does that look in typ­i­cal web analy­sis report­ing? It looks like inef­fec­tive or under-performing paid search and nat­ural search key­words, poorly per­form­ing site con­tent, bad web­site con­ver­sion… and great in-store sales!

That’s the value of a 360° multi-channel view of your cus­tomer:
You gain vis­i­bil­ity into every inter­ac­tion, from the dig­i­tal ini­ti­a­tion to the brick-and-mortar con­clu­sion. And your opti­miza­tion deci­sions can be informed by every aspect of the customer’s experience.

How does this look where the data actu­ally has to come together? Let’s explore…

Dataset Schemas in Insight

You might be very famil­iar with how Web data is typ­i­cally orga­nized. A Vis­i­tor might have one ore many Vis­its, each of which might encom­pass one ore many Page Views or other types of Hits.

As you move to a Customer-centric type of dataset in Insight, your dataset schema also evolves. Rather than just Web Vis­i­tors, there’s a higher-level con­cept of Cus­tomers. Cus­tomers might be Web Vis­i­tors… or store pur­chasers… or peo­ple who have had both types of experiences.

A Cus­tomer might have one or many Expe­ri­ences with you, and each of those Expe­ri­ences might have one or many Events.

We some­times name these lev­els dif­fer­ently, due to vary­ing nomen­cla­tures from orga­ni­za­tion to orga­ni­za­tion. But the high-level con­cept is the same.

If it helps you under­stand dataset schemas a bit more — and the oppor­tu­ni­ties for their diver­sity in Insight — check out my post from a cou­ple of years ago, A Tale of Three Dataset Schemas.

Putting Together the Cus­tomer View

Now that we under­stand how the schema can be cus­tomized and how we’re empow­ered to orga­nize our data how­ever will best suit our analy­sis, we can start to con­sider what data sources we want to include.

Some of the most com­mon data sources in this type of imple­men­ta­tion include Web Data (usu­ally SiteCatalyst-collected data from DataWare­house), Cus­tomer data from your inter­nal data ware­house, cost & impres­sion data from your dis­play ad man­age­ment plat­form, paid search cost & impres­sion data from Search­Cen­ter or another paid search man­age­ment plat­form, send/open/click data from your email ser­vice provider, and more.

We’ve devel­oped the Uni­fied Cus­tomer Process in Adobe Con­sult­ing to deliver this real­ity. It includes the process of dis­cov­er­ing your avail­able data and map­ping it into a uni­fied form, and the tech­ni­cal end of putting it all together in Adobe Insight.

Once we’ve com­bined all of those data­sources, using the keys from each that tie the Customer’s expe­ri­ences together, we’re ready to dive in and get some Insight, Results & Action!

Insight! Results! Action!

Before we orga­nized all of our Cus­tomer data in Insight… back when we were just look­ing at Web data, we may have answered ques­tions like:

Which sites on which I’m run­ning dis­play ads deliv­ered the most Rev­enue & Rev­enue per Vis­i­tor last week?”

Imag­ine the new met­rics we might have avail­able to us with our var­i­ous data points about the Cus­tomer integrated…

  • Ad Spend
  • Ad Impres­sions
  • In-Store Rev­enue
  • In-Store Rev­enue per Customer
  • Total Rev­enue (Web Site + In-Store)
  • True Return on Ad Spend (ROAS)

Our analy­sis can sud­denly look much, much different.

Sud­denly, in one tool, we can see how our spend within a given mar­ket­ing chan­nel impacts our store rev­enue for cus­tomers who inter­acted in that chan­nel, along with our Web rev­enue, the com­bined rev­enue, the return on our investment.

Often, this illu­mi­nates cer­tain efforts — sites, place­ments, cre­atives — that looked great in the online-only view, but aren’t great in a whole-customer view. It also helps us see efforts that looked bad in the online-only view, but are dri­ving real con­ver­sion or high life­time value cus­tomer in the offline world.

And we can select, seg­ment, & fil­ter on any attribute we have in the dataset:

  • Time (when the events happened)
  • Attrib­utes about the cus­tomer from my dataware­house (age, gen­der, etc.)
  • Browse behav­ior (prod­ucts viewed, areas of the site explored)
  • Other mar­ket­ing tac­tics the cus­tomer inter­acted with
  • And more…

As a result, here are some real life opti­miza­tion deci­sions that some of my clients are mak­ing in opti­miz­ing their online marketing

  • Re-adding pre­vi­ously can­celled sites & place­ments that looked “bad” from the Web-only per­spec­tive, but are actu­ally high per­form­ers con­sid­er­ing the offline channel
  • Cut­ting sites & place­ments that looked “great” from the Web-only per­spec­tive, but show lower ROAS than oth­ers con­sid­er­ing the offline channel
  • Opti­miz­ing under-performing place­ments by test­ing new cre­ative that’s work­ing bet­ter at offline con­ver­sion in other sim­i­lar placements
  • Find­ing ways to reduce spend while increas­ing ROAS due to the multi-channel view
  • And, of course, more…

Those are real, high-impact, mar­ket­ing deci­sions & opti­miza­tions hap­pen­ing every day within client orga­ni­za­tions that are employ­ing a 360° multi-channel view of the Customer.

That’s Not All

I’ve shown some exam­ples of visu­al­iza­tions with dis­play ad data here, but don’t for­get about all of the other mar­ket­ing chan­nel data you could include. We have clients actively using Search­Cen­ter (paid search) cost & impres­sion data in Insight, as well as many other mar­ket­ing tac­tics… emails sent & opened, for example.

Still, per­haps you’re not quite ready to tie all of your Cus­tomer together in one place and gain these benefits.

If so, there’s another step you could aim for first, and it entails bring­ing just the data related to what your online place­ments are cost­ing you into Insight and com­bin­ing it with your web­site traf­fic data. I’ll explore that option next week.