In an increasingly multichannel world, it’s essential for marketers to employ attribution modeling to understand how their various marketing channels and customer touchpoints contribute to conversions. However, just 58% of marketers surveyed say they have a working attribution model. Despite all the value attribution modeling can add to your business, it clearly comes with significant challenges that can make it difficult to roll out and maintain.
In this post, we’ll explore a few of those challenges that can prevent marketers from creating and refining their attribution models. From building the right internal infrastructure and streamlining organizational practices to the data siloing and challenges posed by online privacy and tech, we’ll help you understand the underlying issues that can prevent you from reaching your attribution goals!
Inconsistent URL/Media Naming & Tagging
If you’re looking to build out your attribution model for the first time or to improve your existing model to reflect changes in your marketing strategy, then a consistent & scalable web analytics taxonomy for your marketing assets is critical. To accurately report on and understand the way prospects interact with your brand across marketing channels, you need at a minimum URL parameters, aka UTM parameters for Google Analytics, scaled across tactics and channels at the desired level of detail.
For example, in order to derive meaningful and actionable insights, all of your paid media links should contain Source, Medium, and Campaign parameters, ideally following a consistent naming convention. Here are some definitions / suggestions for how to name these 3 across channels:
- Source – Name of the referring website, platform, search engine, newsletter name, or other source. Examples are Facebook and Google.
- Medium – Refers to the type of referrer, e.g. Paid Search, Email, or Social.
- Campaign – Name used to identify a specific product promotion or strategic campaign, e.g. Regional Sales Support, or Back to School Sale. This parameter is critical for data analysis of cross-channel initiatives.
Problems frequently emerge either when multiple teams field different parts of a campaign without centralized coordination or try to change naming conventions while legacy assets are still active in the field. Common issues that we see tend to fall into the following categories:
- Typos and inconsistent spellings, e.g. E-mail & email for the medium value
- Inconsistent cross-channel campaign names, e.g. “2019 Back to School Sale” vs. “Back to School”
- Using the wrong UTM parameter, e.g. using Facebook for the Medium value, instead of Source
Overcoming this challenge can require a lot of effort, but developing a centralized process or a single set of tools, such as a UTM link generator or even a simple project management tool, to control the tagging can eliminate a lot of confusion, making it easier to capture the data you need to properly attribute your results.
Overly Detailed or Rigid Processes
Developing an approach for such tasks as building trackable links and tagging media assets is key, but it’s also possible to overcomplicate the process. Multi-touch marketing attribution analysis requires larger volumes of data than last-touch conversion analysis. Therefore you may not have enough data to conduct cross-channel analytics below the above-mentioned levels of Source/Medium/Campaign. Furthermore, the more levels of detail you require for your URL/UTM generator, the more likely it is to result in non-compliance or errors.
A similar issue can arise when trying to impose a standardized naming convention on different paid media, such as paid search and display. For example, a business promotion doesn’t necessarily fit well into the campaign naming structure of a paid media buy in Google Ads or Facebook. The latter have a pre-existing structure based on the media-buying nuances of the platform. A strategic campaign such as a Back to School sale will likely exist at the Ad Group level for paid media, with the campaign being something like Non Brand – Promotions.
The beauty of these tools is that you have much more flexibility to allow channels to maintain their nuanced naming elements, while also applying higher level elements such as an overarching strategic initiative or campaign. For example you may choose to call the higher level element a “Strategic Campaign” or “Promotion” while still reporting on the lower-level elements such as (Paid Search/Social) Campaign or Ad Group.
Your marketing tech stack probably includes multiple platforms and tools, all of which house some of the data that you need to build your attribution model. For instance, if you’re using attribution tools within Google Ads or Campaign Manager, those models will only be built from the data that’s tracked by those platforms. For example, if you look at a top paths attribution report within Google Ads, this will only include Google Ads clicks, even if users may have clicked on an organic search ad or an email link prior to converting on your website.
To really make attribution work for your business, you’ll need to integrate data across all available digital media platforms, web analytics, and if relevant, offline or backend sales. At a minimum this will help you achieve true “deduped” last touch attribution, while simultaneously preparing an integrated data model for more advanced rules-based and data-driven attribution modeling.
The complexity of the digital ecosystem also throws up some roadblocks to an optimal attribution strategy. To understand which touchpoints and messages influenced the customer journey, you need to be able to identify the customer at every stage of that journey. In the days when desktop browsing was king, marketers could rely on cookie tracking to positively identify a customer at every step of their journey. The proliferation of mobile devices, apps, and the walled-garden ecosystems of major social platforms makes this more challenging. Marketers must rely on signals from a myriad of third-party cookies integrated manually or by an array of tools to ensure that they are following a customer’s full journey from desktop to mobile and through a variety of closed app ecosystems.
Adding to this challenge, the third-party cookie is itself an endangered species. A host of new regulations, including the California Consumer Privacy Act (CCPA) and the European Union’s General Data Protection Regulation (GDPR), seek to curb marketers’ ability to track customers across the web. In recent years, a number of browsers, such as Firefox and Safari, have responded to regulatory pressure and consumer concerns about privacy by rolling out no-tracking options and, more recently, enabling these features by default, meaning customers would have to manually choose to allow active tracking and disable the browser privacy feature. Google recently announced that it plans to phase out third-party cookies across its popular Chrome browser, which represents roughly 68% of the entire market.
Marketers aren’t powerless against this challenge. Many have partnered with publishers and demand-side platforms to acquire more customer insights without tracking customers across the web. Others have worked to build their own trusted ecosystems in which customers opt-in to sharing data in exchange for a better and more personalized experience, creating an even exchange of value that satisfies regulatory requirements and public opinion. More recently, Google announced that it will not leave marketers to twist in the wind. The company will fully abandon third-party cookies by 2022, but it will roll out a suite of browser tools to help marketers balance demands for privacy with the need to enhance personalization. These changes should allow marketers to continue to build and refine their attribution models across channels and devices.
It’s helpful to understand what kind of barriers might stand between your marketing team and a fully realized approach to attribution modeling, but knowledge alone won’t allow you to meet your business goals. Empirical Path’s marketing attribution experts can help you develop the skills and tools you’ll need to overcome the hurdles and build an attribution model that works for your business. Contact us today for a consultation.
how we can help you.