The release of Adobe Ana­lyt­ics ver­sion 1.4 marks a major shift in empow­er­ing mar­keters and mar­ket­ing ana­lysts with pow­er­ful advanced sta­tis­tics and pre­dic­tive ana­lyt­ics capa­bil­i­ties, help­ing move both mar­keters and ana­lysts up in the Ana­lyt­ics Capability-Maturity Self-Assessment Tool. Now, more than ever, mar­ket­ing orga­ni­za­tions can har­ness the power of pre­dic­tive ana­lyt­ics with­out a deep knowl­edge of sta­tis­tics and data min­ing. Lever­ag­ing the new pre­dic­tive ana­lyt­ics fea­tures in Adobe Ana­lyt­ics 1.4 will dras­ti­cally increase the effec­tive­ness of analy­sis and opti­miza­tion, lead­ing to increased suc­cess, decreased costs, and a bet­ter under­stand­ing of your busi­ness and customers.

Among other fea­tures (such as real-time report­ing), Adobe Ana­lyt­ics 1.4 includes pow­er­ful capa­bil­i­ties that eas­ily progress an ana­lyst through the fol­low­ing steps in the pre­dic­tive mar­ket­ing workflow:

  1. Anom­aly Detection
  2. Data Explo­ration
  3. Audi­ence Discovery
  4. Cus­tomer Scoring
  5. Opti­miza­tion

Anom­aly detec­tion lever­ages advanced algo­rithms to auto­mat­i­cally dis­cover and sur­face sta­tis­ti­cally rel­e­vant insights, thus empow­er­ing mar­ket­ing ana­lysts with the abil­ity to make smarter deci­sions, faster than was pre­vi­ously pos­si­ble. For trended data, anom­aly detec­tion iden­ti­fies sta­tis­ti­cal anomalies—that is to say, by sep­a­rat­ing “true” sig­nals from “noise” as data changes over time. This helps ana­lysts unearth anom­alous behav­iors or pat­terns that may have oth­er­wise gone unno­ticed. Anom­aly detec­tion trans­forms ana­lysts into advanced sta­tis­ti­cians by unearthing valu­able oppor­tu­ni­ties ripe for opti­miza­tion.
Avail­abil­ity: Adobe Ana­lyt­ics – Stan­dard (Reports & Ana­lyt­ics, Ana­lyt­ics iOS mobile app, Report Builder)

Data explo­ration assists ana­lysts with dis­cov­er­ing hid­den rela­tion­ships across their data. Data explo­ration con­sists of sta­tis­ti­cal cor­re­la­tions, statistics/mathematical sum­maries, and trend lines. Cor­re­la­tions are use­ful for quan­ti­fy­ing the rela­tion­ship between two vari­ables (mea­sur­ing the degree to which two variable’s move­ments are asso­ci­ated). Statistical/mathematical sum­maries quan­tify the “typ­i­cal” or “aver­age” val­ues as well as the dis­per­sion in a data set. These sum­maries include the mean, max­i­mum, min­i­mum, count, vari­ance, stan­dard devi­a­tion, and interquar­tile range. Data explo­ration not only exposes unknown oppor­tu­ni­ties across data sets but also helps the ana­lyst select impor­tant and/or rel­e­vant met­rics for the next steps.
Avail­abil­ity: Adobe Ana­lyt­ics – Pre­mium (Data Workbench)

Math Functions in Worksheet

Math Func­tions in Worksheet

Audi­ence dis­cov­ery moves beyond sim­ple seg­men­ta­tion toward iden­ti­fy­ing groups (or clus­ters) of sim­i­lar cus­tomers that dif­fer sig­nif­i­cantly from other groups of sim­i­lar cus­tomers. Although seg­men­ta­tion plays an impor­tant role in using “bor­ders” (pre­med­i­tated divi­sions) to sep­a­rate cus­tomers, audi­ence clus­ter­ing finds the “bor­ders” between groups of cus­tomers that pro­duce size­able, mar­ketable, action­able, and valu­able audi­ences for analy­sis and opti­miza­tion.
Avail­abil­ity: Adobe Ana­lyt­ics – Pre­mium (Data Workbench)

Lift Gain Chart

Lift Gain Chart

Cus­tomer scor­ing pre­dicts which cus­tomers are most likely to respond (e.g., churn, pur­chase, click). As a result, the most rel­e­vant audi­ence can be tar­geted. With the release of gains/lift charts in Adobe Ana­lyt­ics 1.4, ana­lysts can pre­dict the impact the scored model will have on their busi­ness. These charts empower ana­lysts with the abil­ity to make con­fi­dent state­ments, such as “by tar­get­ing the top 20 per­cent of the most likely cus­tomers to pur­chase, I can expect a 3x lift in pur­chases, com­pared with tar­get­ing cus­tomers at ran­dom.” Now that the oppor­tu­nity is quan­ti­fied, the audi­ence eas­ily ports into exe­cu­tion chan­nels for tar­get­ing through inte­gra­tions with Adobe Tar­get. This entire process, from anom­aly detec­tion to tar­get­ing, can be com­pleted in as lit­tle as 10 min­utes. Again, with­out a deep knowl­edge of sta­tis­tics.
Avail­abil­ity: Adobe Ana­lyt­ics – Pre­mium (Data Workbench)

Opti­miza­tion is the ulti­mate goal of any ana­lyst work­ing in the field. Using the new advanced sta­tis­tics and pre­dic­tive ana­lyt­ics capa­bil­i­ties built into Adobe Ana­lyt­ics 1.4, ana­lysts can “run lean” by lever­ag­ing time while allow­ing for a more accu­rate fore­casts. Hav­ing an agile ana­lyt­i­cal process means hav­ing the abil­ity to vet data faster and in ways that nor­mally would take much longer. When com­bined with other solu­tions such as Adobe Tar­get, Adobe Ana­lyt­ics 1.4 just makes sense.
Avail­abil­ity: Adobe Ana­lyt­ics – Pre­mium (Data Workbench)