Useful Resources
At Datalicious we regularly produce groundbreaking research to benefit the wider marketing community. In this section you can find a selection of the white-papers and research studies our team published alone or in cooperation with companies such as Facebook, Criteo and Econsultancy.
Facebook: Cookie vs. People-Based Measurement
Massive study analysing >1 billion media touchpoints spread over >80 million purchase paths and >1 million conversions to understand the full impact of using a people-based cross-device attribution model compared to a more traditional cookie-based single-device attribution approach.
Impact of Ad Fraud on Australian Advertising
Display ad fraud can be broken down into two major categories: Ads that were never really served and website visitors that do not really exist. Together, Datalicious estimates that these forms of ad fraud impact nearly 8% of all display ad impressions and account for over $100 million dollars of wasted ad spend.
Facebook: People Based Cross-Device Attribution
Is mobile undervalued as a channel? In this study we used Facebook people based cross-device tracking rather than cookies to understand the true value of mobile ads on desktop conversions and vice versa. The results were presented by our CEO at the Facebook 'Mobile Moves Commerce' event series.
Forrester: Financial Payback For Marketing Attribution
What is the potential ROI from implementing attribution? In their recent report, Forrester outlines the costs, benefits and risks involved as well as estimates that Unified Marketing Impact Analytics (UMIA), otherwise known as marketing attribution, delivers an approximate 15-20% gain in media efficiencies.
Econsultancy: South East Asian Media Mix Index
Our joint survey with Econsultancy on effective marketing mix allocation revealed that Asian marketers are spending their budgets on the wrong channels. Asian executives tend to over-spend on traditional offline channels even though that is not where consumers spend the majority of their time.
Econsultancy: Australian New Zealand Media Mix Index
Even though less pronounced than in the rest of Asia, Australian marketers are still spending their budgets on the wrong channels. Australian executives tend to still over-spend on traditional offline channels even though that is not where consumers tend to spend the majority of their time anymore.
Criteo: Attribution Shows True Impact of Retargeting
Datalicious and Criteo teamed up to demonstrate the true impact that re-targeting delivers to advertisers. Unfortunately, this is not yet completely understood in a digital marketing world still dominated by last-click measurement and cost-per-click remuneration models.
Facebook: Cross-Channel Attribution in Finance
A multi-touch attribution study on the role of Facebook, display and search marketing in financial services. The study, the largest of its kind globally, summarises the results of >700 million media touchpoints across >100 million purchase paths and 75,000 conversions across five months and seven brands.
Econsultancy: Global Media Mix Index Study
A global study completed with Econsultancy comparing media budget allocation to media consumption as a 'rough' proxy for effective allocation of ad spend. In the survey, we asked over 700 marketers on how they allocate their media budgets and then compared that to where consumers actually spend their time.
Econsultancy: State of Media Attribution in Asia
A joint research study completed with Econsultancy, which shows that even though most marketers in the APAC region recognise the importance of accurate marketing measurement, this does not translating into action. Last-click attribution remains the most commonly used model leading to ineffective allocation of spend.
Carnegie: Hidden Markov Model for Attribution
Consumers are exposed to advertising across many channels. As such, a conversion may be the result of a series of ads that were displayed but which of the ads should get credit and how much? The paper by Abhishek, Fader and Hosanagar discusses how Hidden Markov Models (HMM) can be used to answer this question.