CASE STUDIES

A leading e-commerce company forays into quality advertising with MiQ

  • The Challenge

    A leading American multinational corporation and e-commerce company, faces a challenge to retain their market share in the online shopping space versus new entrants and competitors in the E-commerce retail market.

    They developed their own attribution model where more emphasis was given to inventory in view as well as driving quality advertising. As opposed to just post-click as the key metric of campaign success they also wanted to look at the influence on the user within the first seven days since first targeted with an ad.

  • The Solution

    1. numbers-01
    2. The challenge represented by optimising to multiple KPIs at once was tackled by approaching the optimisation from an holistic view. Hence we created multiple combinations of targeting capabilities to layer on the top of our delivery. Automation and dynamic targeting were fully utilised in order to reach the KPIs
    3. ecomm numbers
    4. We ranked users based on their browsing behaviour and categories of site domains they visited and spent most time on. We created customised BYOAs (Build Your Own Audiences), to target highly interested users as well as optimised the campaign.
    5. We also utilised MiQ’s ‘Predict’ capabilities to build and target ‘lookalike’ audience segments based on existing segments provided by the company. Campaign targeting was optimised to yield better results, based on the time, frequency and time lag to conversion of the users
    6. numbers-03
    7. We pooled users into groups based on the domains they visited including automotive, fashion and electronics and specifically targeted the domains that drove the maximum revenue with more bids. Keeping in mind the importance of ‘inventory’ in digital advertising, we created blacklists of non-performing domains, which we updated every week based on performance. We also created dynamic DPIs to exclude all the inventory which weren’t performing well in the past 2-3 days for the given number of impressions served and conversions obtained
    8. numbers-02
    9. In order to test strategies and creative sizes, the campaign setup was split into various placements based on the creative size and we observed which performed better.
    10. Creative optimisation was updated daily to focus on the best creative size that would have been shown for the longest time

     

  • The Result

    • 700% increase in delivery from May, 2015 to October, 2016, with a 70% increase in viewability (37% – 63%)
    • 265% increase in CVR due to Predict