Discover 3 Free Trial - Retail Quick Wins from Adobe Consulting

Over the com­ing weeks, the retail indus­try experts in Adobe Con­sult­ing will share a series of analy­sis quick wins for retail­ers, using Adobe Dis­cover 3. For a lim­ited time, Adobe Site­Cat­a­lyst 15 clients can inquire with their account team and ask to take part in a free trial of Adobe Dis­cover. We’ve made it eas­ier than ever to try Dis­cover, and we’re show­ing some great Dis­cover analy­sis oppor­tu­ni­ties spe­cific to the retail indus­try. For more infor­ma­tion and to request trial access, con­tact your account man­ager or account executive.

Adobe Dis­cover — Retail Quick Win #4
Visitor-Level Fall­out Analysis

One of the most excit­ing fea­tures for retail­ers, in my opin­ion, with the release of Dis­cover 3 is the addi­tion of a ‘Vis­i­tor’ view to sev­eral reports and met­rics. We’ll take a look at some of the power made avail­able by Dis­cover 3 for unique data analy­sis you can do, espe­cially if you are new to Dis­cover or new to Dis­cover 3. Let’s look at a few specifics in how to apply some of these excit­ing new fea­tures of Dis­cover 3.

Per­form Fall­out Analy­sis for Sequences of Actions Across Visits

With the addi­tion of seg­men­ta­tion in Site­Cat­a­lyst 15, we apply seg­ments to your fall­out reports, mak­ing it pos­si­ble to iden­tify the fall­out paths for users who come to your site from a given mar­ket­ing chan­nel. Dis­cover 3 takes this to the next level to allow you to incor­po­rate any num­ber of dimen­sion and events directly into the fall-out and look at that across vis­its. Let’s look at an example.

We want to bet­ter under­stand, at a high level, the flow of our users who sign-up for an email, come back to the site from an email, add some­thing to cart and even­tu­ally pur­chase. By nature, our ques­tion is going to span vis­its and a tra­di­tional Site­Cat­a­lyst or pre-D3 fall-out report won’t cut it.

But can’t we get there with a vis­i­tor seg­ment?” we might ask? While we could cre­ate a vis­i­tor seg­ment which defined vis­i­tors who signed up for an email, came to the site from email, added a prod­uct to cart and then pur­chased, the impor­tant ele­ment miss­ing from the seg­ment solu­tion is sequence. While your seg­ment would include the group we’re look­ing for, it could also include indi­vid­u­als who pur­chased or added to cart before sign­ing up for an email.

To build our fall-out report we sim­ply drag over the events or dimen­sions (eVars or props) we want to include. We can then con­fig­ure the dimen­sions for spe­cific val­ues, such as ‘Email’. An exam­ple of what our fall­out report could look like is shown in Fall­out Dia­gram 1.1:

Fall­out Dia­gram 1.1: Email Signup Fall­out Sum­mary Report


The beauty of the flex­i­bil­ity here is we’re not lim­ited to just pages or just events; we can mix and match any com­bi­na­tion of inter­ac­tions to build very spe­cific analy­sis use cases, either at the visit or vis­i­tor level.

We can also group val­ues of dimen­sions into the fall­out step. For exam­ple, if I wanted to under­stand orders for vis­i­tors who had come from either Dis­play or Social Net­works, I could group them into a sin­gle fall­out check­point, as shown in Fall­out Dia­gram 1.2 and 1.3.

Fall­out Dia­gram 1.2: Dis­play OR Social Net­works Last Touch Chan­nel — Fall­out Check­point Builder

Fall­out Dia­gram 1.3: Dis­play OR Social Net­works Last Touch Chan­nel Fall­out Sum­mary Report

Per­form Fall­out Analy­sis for Sequences of Mar­ket­ing Chan­nel Inter­ac­tions Across Visits

Let’s dig in fur­ther and inves­ti­gate the fall­out between cer­tain chan­nels. We’ll start by look­ing at all vis­its for all vis­i­tors who came in from paid search, but exclude those vis­its where the final touch before con­ver­sion was Paid Search. We want to remove some of the noise from those vis­its where some­one came in and con­verted from Paid Search. An exam­ple of this seg­ment is shown in Fall­out Dia­gram 2.1.  This allows us to zero in on those vis­i­tors who came to the site from paid search, but didn’t con­vert on that visit.

Fall­out Dia­gram 2.1: Paid Search Vis­i­tors Segment

We then apply this seg­ment to our chan­nels report to under­stand what chan­nels users even­tu­ally ordered from when they had also vis­ited from paid search. This gives us the ini­tial infor­ma­tion we want to be able to com­pare the rela­tion­ship of our other chan­nels to paid search to get a sense of which we’d like to dig fur­ther into.

Fall­out Dia­gram 2.2: Last Touch Chan­nel Report with Paid Search Vis­i­tors Seg­ment Applied


We notice from our exam­ple that Dis­play and Com­par­i­son Shop­ping have the high­est num­ber of vis­its and orders, so let’s dig in fur­ther to each of these channels.

We can incor­po­rate each of these chan­nels into their own fall­out report and then com­pare how they per­form against each other. In so doing, we’re able to see which one per­forms bet­ter with vis­i­tors who had pre­vi­ously clicked through on a paid search ad and begin to under­stand the rela­tion­ship between paid search and these two other channels.

To ensure we are work­ing with mutu­ally exclu­sive pop­u­la­tions (to help remove any noise we might expe­ri­ence from users who used both Dis­play and Com­par­i­son Shop­ping) we’ll cre­ate a sim­ple exclude vis­i­tor seg­ment for each fall­out report. An exam­ple of the seg­ment for exclud­ing Com­par­i­son Shop­ping from the Dis­play Ad fall­out report is show­ing in Fall­out Dia­gram 2.3.

Fall­out Dia­gram 2.3: Vis­i­tor Exclude Com­par­i­son Shop­ping Engine Segment


After apply­ing each seg­ment to a unique report, we then build our fall­out reports and make sure we have the ‘vis­i­tor’ option selected. I like to include all vis­i­tors to get a sense of how the pop­u­la­tions com­pare to each other. Addi­tion­ally, we can also spec­ify on the last chan­nel, that this chan­nel was the chan­nel where the order took place. We do that by select­ing the ‘Order’ event from the Check­point Edi­tor as shown in Fall­out Dia­gram 2.4.

Fall­out Dia­gram 2.4: Dis­play with Orders Event Fall­out Check­point Builder

As we now exam­ine our fall­outs, we’re able to make some inter­est­ing observations.

Fall­out Dia­gram 2.5: Paid Search Fall­out with Last Touch Dis­play Sum­mary Report

Fall­out Dia­gram 2.6: Paid Search Fall­out with Last Touch Com­par­i­son Shop­ping Engines Sum­mary Report

First, we notice that the pop­u­la­tion we’re deal­ing with for Com­par­i­son Shop­ping is about ¾ of the pop­u­la­tion of Dis­play. Imme­di­ately we start to under­stand brows­ing behav­ior for vis­i­tors who have come through from Paid Search may be less likely to come back and con­vert from Com­par­i­son Shop­ping than Dis­play. This could be impacted by the over­all mar­ket­ing strat­egy or asso­ci­ated mes­sag­ing of each dis­tinct chan­nel. Addi­tion­ally, we observe that users com­ing back from Dis­play con­vert more than three times as high Com­par­i­son Shop­ping. Again, we may want to exam­ine what our over­all strat­egy or mes­sag­ing is in rela­tion to Com­par­i­son Shop­ping. Armed with this infor­ma­tion about the rela­tion­ship between Paid Search and Dis­play, we could adjust our mes­sag­ing for Display.

Final Thoughts

For retail­ers, fall­out analy­sis in Site­Cat­a­lyst has been one of the ‘Go To’ analy­ses for find­ing hid­den mon­e­tize­able oppor­tu­ni­ties. With the addi­tion of seg­men­ta­tion in Site­Cat­a­lyst 15, ana­lysts have been given more flex­i­bil­ity to quickly answer ques­tions about how users inter­act with aspects of their site dur­ing a visit. Now, with Dis­cover 3, retail­ers can expand their ques­tions to include inter­ac­tions occur­ring across mul­ti­ple vis­its, answer­ing ques­tions they couldn’t before. If you haven’t dug into this excit­ing fea­ture, I’d encour­age you to carve out some time over the next week to try it out.

Derek Tan­gren is a Prin­ci­pal Con­sul­tant and Team Lead in Adobe Con­sult­ing, focused on dig­i­tal strat­egy, ana­lyt­ics & opti­miza­tion for retail & travel clients. He tweets at @derektangren.

If you’re an online or cross-channel retailer using Adobe Site­Cat­a­lyst 15, you should try these Retail Quick Wins in Adobe Dis­cover. We’ve made it eas­ier than ever to expe­ri­ence a free trial of Dis­cover. For more infor­ma­tion and to request trial access, con­tact your account man­ager or account executive.

0 comments