GUEST COMMENT Understanding attribution to improve entire marketing strategies
by Adit Abhyankar
The science of attribution is critical to the success of today’s multichannel marketing campaigns. It gives brands a clear perspective of which touchpoints worked most effectively to drive conversions, both online and offline, while core attributes of each touchpoint – such as keyword, placement and ad size – can be taken into account. This allows marketers to assess the outcome, adjust their tactics, and maximise ROI.
Retail is a particularly ROI-driven sector, so understanding marketing effectiveness across different audience segments and channels is key. Further to this, retailers often rely on offers and coupons to drive sales, which poses the challenge of understanding the impact of the offer relative to that of the choice of media or creative. This understanding is crucial to ensure that retailers do not cannibalise sales that would usually occur in the absence of an offer, which has obvious implications for ROI. The retail industry focuses heavily on repeat purchases so media choices need to be aligned and optimised to ensure lifetime value.
It’s true that attribution is an advanced marketing diagnostic, but it’s not unattainable to the internet retailer. Nor is there any real mystery to understanding its immense value or finding holistic software solutions to help retail marketers improve their cross-channel performance. In essence, attribution helps answer two key questions:
• Does my online stimulation drive offline response?
• Does my offline stimulation drive my online conversions?
In pursuing the first question, it’s important to ascertain how many offline converters were exposed to a brand’s digital marketing efforts and subsequently, the impact of those digital efforts relative to the impact of brand equity and other outside stimulation, including offline media, seasonality, couponing and special offers. Good attribution systems can build robust scientific models that capitalise on data available at the user level, as well as data that is only available at the aggregate level.
A key focus once these questions have been answered is to assess the key decisions a brand is looking to make, including a review of digital spend. If a business realised the true cost per action (CPA) was actually 20% less than they initially thought because it drives offline conversions, would they spend more on the channel as a whole? And on which channels should they spend more, and when? Does digital media have a higher or lower impact during peak season?
In terms of the second question, it would be relevant to understand whether a company’s digital media is being over-credited for conversions. The answer to this can be found through an improved understanding of what is incremental about digital advertising relative to what would be delivered as a function of offline media and brand equity. Good attribution systems not only credit online conversions to media, but also help unlock the fundamental question of incrementality. Furthermore, attribution should help articulate exactly which offline stimulation and online response channels work best together.
Brands tend to struggle with multichannel measurement for organisational and political reasons, along with a skewed perception of complexity. Most companies organise their marketing by channels – it is how the team is organised, how goals are set and how they are compensated. These channels then map to agency counterparts that are also structured this way, so a company can question how many conversions they are gaining from search or from SEO, and so on. Attribution knocks down these walls because it is clear how a conversion was a product of the efforts of multiple groups. But how is partial credit given to all players? This internal process of reversing this way of thinking is a key reason for friction and resistance.
Related to the organisational problem is the issue of the status quo and vested interests in protecting this position. Often it is those running the most successful channels (or those perceived to be most successful) who end up gaining the most budget, which in turn pushes them into more senior roles. When the silos are broken down and there is a better understanding of cross-channel interactions, the perception of what works and what doesn’t changes, and for some, it can be quite a dramatic shift. It takes a bold company to shake up long-held practices and sometimes the political friction of getting buy-in from senior people can be too much.
Then there is the perception of complexity – it can appear hugely daunting to replace something as fundamental as a marketing measurement system, but while it is a big endeavor, it can be broken into manageable steps:Step one:
Does cross-channel attribution make sense for your business? Step two:
If the method of measurement was changed, what should replace it?Step three:
What would be the impact of such a change and what would it be worth to your business? Step four:
How can trust be built in any new measurement system and how can future decisions be improved upon? Step five:
How can the organisation be changed and senior level buy-in be achieved? Step six:
How can attribution be incorporated into media buying workflows and into better understanding of customers on an ongoing basis?
It’s key to remember that marketing attribution is a relative, not absolute science. There is no nirvana state where an attribution solution can absolutely predict human behaviour down to the individual level 100% of the time. However, a robust science-based attribution platform should be able to seamlessly help a business translate cross-channel insights into specific recommendations for action.
Mature attribution vendors can help marketers throughout this entire journey. The key is to focus on the prize at the end – a vastly better marketing organisation that can both scale and improve the efficiency with which it drives revenue growth – and not let the naysayers get you down.Adit Abhyankar is executive director, Visual IQ.