\n<\/p>\n
Attribution analysis enables\u00a0marketeers to understand\u00a0how each marketing channel\u00a0has\u00a0contributed to the sales effort.\u00a0Attribution analysis is a crucial tool for\u00a0any\u00a0analysts\u00a0who want to understand\u00a0their\u00a0customers and optimise\u00a0the ROI of their\u00a0marketing campaigns.\u00a0In this article we\u2019ll look at how it works and how\u00a0to use it correctly to\u00a0leverage the maximum value.\u00a0<\/p>\n
Simply put, Attribution is\u00a0a\u00a0method used in Digital Marketing\u00a0to identify how much each marketing channel\u00a0has\u00a0contributed to your sales efforts.\u00a0It seems quite straightforward on the surface,\u00a0but\u00a0when you\u00a0dig into the topic, it can\u00a0rapidly become\u00a0complex. However, with the right processes and methodology, a properly functioning\u00a0Attribution analysis can\u00a0significantly\u00a0help\u00a0you to\u00a0optimise\u00a0your marketing ROI.\u00a0<\/p>\n
Today, online consumers\u00a0interact with brands in\u00a0increasingly\u00a0divers ways. Technology provides\u00a0an\u00a0ever expanding\u00a0number of\u00a0channels\u00a0to push\u00a0the right brand message\u00a0to the right user\u00a0when\u00a0they\u00a0are\u00a0most receptive\u00a0\u2013\u00a0and\u00a0on\u00a0an increasing number of devices.\u00a0When a user converts, it is\u00a0also\u00a0more and more\u00a0complex to\u00a0understand\u00a0which interactions between a brand and a user\u00a0have\u00a0truly\u00a0had an impact.\u00a0\u00a0\u00a0<\/p>\n
Recreating\u00a0consumer\u00a0journeys and deciphering this\u00a0type of information\u00a0can be extremely\u00a0complicated.\u00a0Aside from data quality issues and the reliability of\u00a0third-party\u00a0cookies,\u00a0consumer journey datasets are quite complex.\u00a0\u00a0<\/p>\n
It can get tricky, quickly,\u00a0and I\u2019m not even going to\u00a0mention\u00a0the potential\u00a0biases\u00a0of\u00a0third-party\u00a0Attribution\u00a0providers\u00a0when\u00a0they have stakes\u00a0in\u00a0various\u00a0ad networks.\u00a0<\/p>\n
Nevertheless,\u00a0when approaching this topic with a scientific and methodological approach\u00a0\u2013\u00a0and\u00a0by removing the external factors you can control\u00a0\u2013\u00a0it is entirely possible\u00a0to get a clear picture of what is happening behind the scenes. Combined\u00a0with\u00a0computational power\u00a0and some economies of scale,\u00a0an\u00a0accurate\u00a0attribution analysis\u00a0with the correct interpretation\u00a0can provide you with the right answers and insights. This will\u00a0ultimately help you\u00a0optimise\u00a0the\u00a0ROI\u00a0for your marketing initiatives.\u00a0<\/p>\n
Before going into what attribution can do for you,\u00a0here\u2019s a definition:\u00a0\u00a0<\/p>\n
Attribution in online marketing is the process\u00a0that\u00a0credits conversions to the correct marketing activity,\u00a0or combination of activities.\u202fAttribution analysis provides an in-depth\u00a0study\u00a0of marketing activity\u00a0performance over a chosen analysis period,\u00a0against an attribution model of your\u00a0choice.\u00a0<\/p>\n
The first step to\u00a0Attribution\u00a0is to recreate\u00a0customer\u00a0journeys\u00a0\u2013 and\u00a0for this you need\u00a0a 360\u00b0\u00a0view of\u00a0your\u00a0consumers. You need to measure\u00a0all the interactions\u00a0between\u00a0consumers\u00a0and your brand\u00a0across multiple devices and\u00a0across\u00a0all\u00a0your various\u00a0platforms,\u00a0and then\u00a0organise\u00a0them\u00a0in\u00a0chronological order.\u00a0<\/p>\n
Once\u00a0you understand\u00a0the journeys taken by users from brand discovery\u00a0and awareness,\u00a0through\u00a0to\u00a0consideration-to-purchase,\u00a0you\u00a0can apply\u00a0a range of\u00a0attribution\u00a0models. This\u00a0attaches\u00a0weight to the\u00a0different interactions\u00a0based on\u00a0their contribution\u00a0towards\u00a0each step\u00a0consumer journey.\u00a0\u00a0<\/p>\n
There are\u00a0two\u00a0broad categories of attribution models each with their\u00a0own\u00a0advantages and disadvantages\u00a0\u2013\u00a0Rule-based attribution models\u00a0and\u00a0Algorithmic\u00a0models.\u00a0<\/p>\n
Rule-based attribution models<\/strong>\u00a0have the advantage of being cheap, easy to\u00a0implement\u00a0and are completely unbiased.\u00a0This category can\u00a0be divided into\u00a0the following models:\u00a0First\u00a0click,\u00a0Last\u00a0click,\u00a0Linear,\u00a0Time-decay and U-shaped.\u00a0<\/p>\n There is also a\u00a0subset\u00a0of rule-based attribution models:\u00a0cust<\/strong>om<\/strong>\u00a0<\/strong>rule-based<\/strong>\u00a0models<\/strong>.\u00a0Here the analyst defines\u00a0which weights to apply to each interaction based on how it is positioned in the\u00a0consumer journey. This method has the advantage\u00a0of tailoring rule-based models to the business\u2019\u00a0strategy.\u00a0However, the\u00a0model\u00a0is less\u00a0impartial\u00a0and takes\u00a0some\u00a0prior knowledge of\u00a0consumer journeys to setup.\u00a0<\/p>\n This category of models uses\u00a0advanced data science and machine learning techniques\u00a0to\u00a0identify\u00a0tailored weights to apply\u00a0to each interaction that led to a sale,\u00a0based\u00a0on\u00a0a\u00a0wide\u00a0variety of factors.\u00a0These models\u00a0provide an unbiased approach to attribute credit for sales\u00a0and\u00a0can\u00a0provide\u00a0considerable\u00a0results\u00a0with economies of scale.\u00a0However,\u00a0there are a few\u00a0drawbacks\u2026\u00a0These types of models require large amounts of data,\u00a0time,\u00a0and development\u00a0for\u00a0them\u00a0to be accurate. In addition, they require advanced knowledge in data science to set up and can prove to be\u00a0a\u00a0costly\u00a0investment\u00a0to\u00a0develop and maintain,\u00a0hindering their main goal: aiding in optimising marketing ROI.\u00a0<\/p>\n Now that we know how\u00a0attribution\u00a0works\u00a0as well as\u00a0the\u00a0pros and cons\u00a0of the various attribution models,\u00a0you\u00a0can\u00a0focus on which one you should apply to your business.\u00a0<\/p>\n There is no\u00a0right or wrong\u00a0attribution model;\u00a0they will all provide\u00a0a\u00a0valuable\u00a0perspective\u00a0to\u00a0understand consumer behaviour and\u00a0optimise\u00a0marketing ROI.\u00a0The key is in selecting the\u00a0optimum models\u00a0\u2013\u00a0those\u00a0that\u00a0most accurately\u00a0reflect\u00a0your business strategy\u2026\u00a0<\/p>\n Are you\u00a0currently\u00a0focused on user acquisition\u00a0and brand awareness?\u00a0\u00a0 \u00a0 As well as\u00a0using\u00a0multiple perspectives\u00a0and benchmarking\u00a0to\u00a0analyse\u00a0your\u00a0progression,\u00a0it\u2019s\u00a0also\u00a0important to include\u00a0human validation to\u00a0contextualise\u00a0the results\u00a0and ensure your data makes sense. This also enables you\u00a0to understand the value of each\u00a0touchpoint\u00a0and\u00a0how it fits in with the other\u00a0interactions with your customers.\u00a0<\/p>\n\n
Algorithmic Attribution Models\u00a0<\/strong><\/h3>\n
How do you define your Attribution strategy?\u00a0<\/strong><\/h2>\n
<\/strong>\u2192\u00a0Analyse your data via the lens of a first click\u00a0model.\u00a0\u00a0
\u00a0
Are you having\u00a0trouble converting\u00a0potential customers\u00a0after they are exposed to your brand?<\/strong>
\u2192\u00a0Look at using a last\u00a0click\u00a0model.\u00a0<\/strong>\u00a0
\u00a0
Do you have\u00a0long consumer journeys\u00a0from discovery\u00a0to purchase\u00a0and every touchpoint is critical?
<\/strong>\u2192\u00a0Look\u00a0at using a linear\u00a0or time-decay\u00a0attribution model.\u00a0<\/p>\n
Once you have\u00a0identified\u00a0the\u00a0relevant\u00a0attribution models\u00a0to your business,\u00a0you need to\u00a0stick with them and benchmark performance\u00a0over time\u00a0to measure improvements in your\u00a0optimisations.\u00a0<\/p>\n