In a recent blog post, we announced the new, free Adobe Tar­get Value Nav­i­ga­tor, a unique tool that helps com­pa­nies dis­cover the most impact­ful opti­miza­tion meth­ods to drive con­ver­sion rate lift on var­i­ous pages through­out their web­sites. In this post, we’ll take a look at the math behind the tool and pro­vide insight into the results that users might see.

In devel­op­ing this tool, our engi­neers worked closely with Adobe Con­sult­ing Ser­vices to put together typ­i­cal median lift ranges our con­sul­tants have observed over nearly a decade of engage­ments with hun­dreds of orga­ni­za­tions of all sizes. The algo­rithms behind the tool are built on top of unique lift ranges for each com­pany objec­tive and web­site page. Where a com­pany will fall in those ranges depends on the website’s total con­ver­sions, the opti­miza­tion tac­tics deployed and the resources avail­able within the orga­ni­za­tion. This tool asks these impor­tant ques­tions and returns the poten­tial lifts within the lift range that our con­sul­tants typ­i­cally see.

Step 1 of the Value Nav­i­ga­tor (see screen shot below) asks users to spec­ify their ulti­mate opti­miza­tion goal (for exam­ple, increas­ing rev­enue per vis­i­tor) and rate their com­pany in terms of opti­miza­tion sophis­ti­ca­tion (level of resources, exec­u­tive involve­ment, etc.). Cer­tain goals are more likely to impact a website’s con­ver­sion rate than oth­ers. For instance, a goal of increas­ing rev­enue per vis­i­tor is likely to gen­er­ate more lift than sim­ply increas­ing over­all web­site traf­fic. This step is the ulti­mate start­ing point for the algo­rithm. It is what we used to decide which range they should start in. Also, more sophis­ti­cated orga­ni­za­tions con­sis­tently place higher on the con­ver­sion rate lift curve—the more test­ing you do, the more lift you’ll see. For the algo­rithm, this means the more sophis­ti­cated the orga­ni­za­tion, the fur­ther up the lift scale it will slide.








Con­ver­sion rate lift ranges are also affected by the spe­cific tar­get­ing meth­ods a com­pany intends to imple­ment, such as auto­mated deci­sion­ing and tar­get­ing, mul­ti­vari­ate test­ing, or rec­om­men­da­tions and cross-selling. This data is col­lected in Step 2 (see screen shot below). Sim­ply put, the more meth­ods you use, the higher con­ver­sion rate you’ll see. We used tar­get­ing meth­ods as our stan­dard devi­a­tion for the algo­rithm. The more meth­ods a com­pany imple­ments, the larger swing in con­ver­sion rate it will see.







The pages on a web­site on which tar­get­ing will be imple­mented is a crit­i­cal fac­tor in the equa­tion, because some pages will have a big­ger impact on your con­ver­sion rate. For exam­ple, dif­fer­ent opti­miza­tion meth­ods tend to work bet­ter on a prod­uct page ver­sus a home page or land­ing page. Our equa­tion employs a mul­ti­plier based on the poten­tial impact of tar­get­ing on a par­tic­u­lar page type (3x for a prod­uct page, 2x for a home page, etc.).







Once all the inputs are gath­ered (based on tar­get­ing objec­tives, meth­ods and sophis­ti­ca­tion), the tool finds and returns the appro­pri­ate lift range. In the accom­pa­ny­ing screen shot of Step 3, we see that based on the data this user has entered, our equa­tion indi­cates that using auto­mated deci­sion­ing and tar­get­ing on prod­uct pages, land­ing pages, and the home page has the poten­tial to increase aver­age con­ver­sion rate lift by 111%.

The Adobe Tar­get Value Nav­i­ga­tor has been very pop­u­lar since its launch, and many orga­ni­za­tions have used the tool to com­pare lift ranges and iden­tify the opti­miza­tion meth­ods that are likely to be most effec­tive. A com­mon take­away is that although A/B test­ing is some­times a great choice for a par­tic­u­lar goal or page, there are many other capa­bil­i­ties of Adobe Tar­get that can yield higher lift ranges in cer­tain sce­nar­ios. There are many tar­get­ing meth­ods, and this tool will help you decide which opti­miza­tion tech­niques are best for each of your pages.