Kamal Chembath Blog - Advertising Axis - Digital Marketing and Analytics

Testing further down the funnel requires sophistication and a mature, practiced testing approach. While there’s still much room for improvement, WhichTestWon’s 2015 State of Online Testing report shows the testing industry is heading in the right direction.  This year, more testers than ever before ran highly sophisticated tests. They also used more advanced testing and measurement techniques. Curious what we mean by “sophisticated” tests and techniques? Try your luck at the next few questions to find out.

Harnessing the power of ‘thank you’ is important. It’s also often overlooked.  Testers tend to get so wrapped up generating leads and sales, they forget to test the receipt / thank you page. But testing this last step in the sales funnel offers a golden opportunity. That’s because roughly 40% of traffic landing on your thank you page will convert again for something else.  Take advantage of this high conversion rate by testing ways to bring your customers back – again and again. When testing thank  you pages, make sure to think about your audience. Find ways to cater to both first-time and return visitors. Both have different needs. Consider testing different offers for each group to keep them coming back. Test  your thank you emails, too. Did you know new customer thank you emails have the highest open rate of any message you’ll ever send? That’s an insane amount of attention you’ll never get from your customers again. Test ways to extend conversion...

Creating exit offers makes it difficult for your customers to leave without a conversion of some type. Exit offers might come in the form of an animated pop-up, a special offer, or something else that keeps visitors on your site longer. Yes, creating exit offers can create annoyance. But, when done correctly, they can also be highly beneficial. In this award-winning real-life test, using  an exit offer increased conversions an astounding 473%! Use exit offers to your advantage. Test new techniques and ways to create it.  Whatever you do, don’t let prospective customers easily leave without trying to get at least a micro-conversion from them.​

Most customers don’t want to take the time to create / register an account. Or, if they’re returning customers, they might not remember the email or password used to register. Consider testing removing, or at least simplifying, the registration process.  HOWEVER. . . if you do NEED visitors to register, test for savvy ways to make the registration process more appealing. In this real-life test, The Honest Company tested a takeover-style registration form versus a pop-up registration form. The more appealing takeover-style registration form (Version A) increased membership registrations 15.4% over the homepage/overlay version.  

Research shows, on average, about 70% of shoppers will abandon your site at checkout.  This statistic means for every 100 shoppers adding something to their cart, only about 30 will actually go through with the final purchase. The other 70 will leave without buying anything! Shopping cart abandonment is one of the worst curses of ecommerce. The good news is that there is a remedy: testing! If you test ways to improve your checkout process, you can hugely improve your bottom line and online sales.

Flipkart uses an s_code version lesser than h.25.3 code and fails to take advantage of first party cookie (s_fid) set by js code while using a third party cookie. Paying huge sum to Adobe and failing to take advantage of this fall back method :-(  in an online world which experiences a higher rejection rate for third party cookies ????? Adobe used user agent and IP address to track the user while s_vi cannot be SET in version lesses than H.23.5 which only means many people in a single network is treated as one visitor :-( Why on earth do you want to track your visitor statistics in such a way!!!!

Preserving the required metrics/dimensions in a certain order while creating reports using Report Builder, and saving the templates by naming it properly is the key to build automated dashboards. Keep the raw the data in one sheet (don't forget to set the preserved order), and the formatted one in other (just apply the formula once - connect the data with charts). Every time you refresh the report with a different date range or different RSIDs due to preserved order set in the raw data sheet, the formulas applied in the formatted sheet connecting the raw data works superbly.