Adobe Analytics: Best Practices for Implementation

• Digital Analytics

Many digital analytics professionals are forced to ask this question again and again over the course of their career. Are we following the right methodologies and practices in order to get the maximum return from our Adobe Marketing Cloud implementation?

Simply asking yourself this question is an important step towards building the appropriate digital analytics practice within your organization, so bravo. By following a set of proven best practices, you can influence your entire organization to become more data-driven and less gut-driven (which unfortunately is still the norm in many organizations). Here are some best practices from Blueocean Market Intelligence that we have collected over the years that can help professionals embark on a successful implementation journey:

1. Robust foundation: Get your basics right. Unless business owners, partners, agencies and other stakeholders trust the data – no one would trust our recommendations in this area. Start early and start right. It is important to take the time to build a robust foundation.
2. Focus on documentation: Inherent to human nature, we tend to ignore one of the most important facets of implementation – creating, reviewing and managing the correct set of documents. Often organizations make a big mistake by not concentrating on technical specifications document, playbook and the solution design reference (SDR) documentation. There is also lack of intent to own and maintain important use cases in form of tech spec or SDR, which creates a complete lack of history for new folks joining the team.
3. Continuous stakeholder roadshows – Perceived as a complex work stream, implementation is looked down upon as an area of disinterest by most business users. Of course, with the advent of tag technologies, it is changing. However, it is important to have continuous roadshows with stakeholders so the preconceived notions are removed.
4. Creating the right benchmarking mechanism – Painful but true, traditional business owners just shut down Adobe Analytics data sets as the numbers do not match their internal system. Time to take a reality check. Two systems will never match – aim to have a consistent variance between the two systems and a goal of 2-5% during the implementation stage itself.
5. Get in early – Get involved in the projects as early as possible. Or you may end up missing the bus at a later stage and it will be difficult to understand the core concepts.
6. Developing a request management process – Even during the post implementation phase, it is important to keep track of all key requests.
7. Continuously measure metrics and create short-term and long-term roadmaps – important to keep a working and ready to review feasible plan.

To add on, here are some technical recommendations that can help you start early and fast:

1. Make it scalable: Create the right JavaScript and data layer to facilitate integration of tag manager with your page source code. This is important since if it is not done properly, it can create scalability issues with your website codes at a later stage.
2. One common JavaScript file – Blueocean Market Intelligence has learned from our engagements that it is better to maintain one common JavaScript file for ensuring minimal conflicts with other page source codes.
3. Check the default plug-ins first – The APIs and other plug-ins as part of Adobe Analytics JavaScript need to be checked on staging before being pushed into production. Common errors have created issues of breakages with the analytics JavaScript, impacting the page tracking and metrics tracking on the web pages.
4. tl () and s. t () tracking: Both functions are used for different purposes altogether. The former is a page view function and the latter is a link tracking function. Blueocean Market Intelligence has seen organizations use s.t () for tracking clicks on an element, which may lead to double counting of page view metrics. Associated variables would need to be populated again leading to inconsistent and highly inaccurate data across key reports.
5. Restrict usage of variables – We advise that you don’t use up all of your available custom variables. You can use out of the box solutions to derive intelligence. For example, the popular SAINT tool for managing campaign metadata rather than using multiple variables to manage the same.
6. Friendly page names – Painful yes, but it is an implementation must-have. We recommend to all of our clients to define standard naming conventions for all pages and follow them across all the web pages.

By following just a few of these best practices, you should see more progress towards the effective adoption of Adobe Marketing Cloud. The key with modules, like Target and Audience Manager, and their usage, relies heavily on putting the right implementation strategies in place from the beginning.