The Divide
Great, we have a good idea of who should be on the various teams and tracks and what everyone should be doing. We can formally break this across a RACI chart (Responsible, Accountable, Consulted, Informed) and include it in our Project Plan. We can also decide when and how these teams should meet in our Communication Plan, how the analytics function should be structured within the company, and when various stakeholders should be consulted. Most importantly, we can assign proper accountability.
Let’s go back to our Marketing Stack. Analytics sits firmly at the foundation of any marketing campaign or website change. It’s second only to properly maintaining our infrastructure. Another reason our data is often F**ked is that we don’t maintain it, which is why accountability and responsibility are essential. If we continually make changes to our website and apps without building analytics as a recurring task, we will continue to break it.
Product Development Cycle
A Product development cycle should be like a CRO cycle:
Source McKtui Consulting, based on Eisenberg, B., & Eisenberg, J. (2005). Call to Action: Secret Formulas to Improve Online Results. Thomas Nelson.
Often, it’s not because there are so many things people want to improve or fix. CRO often fails to deliver or show clear returns on investment because bad UX is death by a thousand cuts rather than some big profitable change that can be identified to fix everything. Actual continuous improvement requires continuous measurement, analysis, optimising, and testing. And faith.
To understand your data and move to impactful insight and, finally, wisdom, those accountable for tracking must know you’re making changes to the website or app. They need to be informed or consulted, and measurement plans must be updated. If you want them to support you in analysis and understanding whether your improvements were actual improvements, then the analyst also needs to have been part of the discussion, implementation, and results to have enough information to sift through the data. They need some knowledge of what has transpired and why, to give you an understanding of whether it worked.
Constantly working through a wishlist of changes from multiple points of view and stakeholders is like forever chasing your tail, and when you succeed, how confident can you be that it was down to your actions and not just a fluke catch? This is where a smaller organisation can have more success. There are fewer teams, fewer variables, and fewer other things happening. They’re often closer to the problem and the possibility of a sound judgement based on gut feeling, but with a corporation and big brands, just proving causation can be impossible, and a sounder method is needed.
Whether they are internal or external, bring your analytics team out of the cold and keep them up to date. Avoid kneejerk actions and reactions, and be methodical. A lucky win now is not worth the sustainable small gains from genuine, reliable continuous improvement. If you do this, you can’t lose because every loss will contribute to the overall learning and win. This is a true cause to be championed and supported.
I worked with a company that, for a period, worked hard to understand its sales funnel. They implemented detailed Tier 3 tracking for every field, maintained that tracking, and included analysts in their product discussions. Over time, conversion rates increased, and hundreds of thousands of dollars were plugged in their once-leaky funnel. Then leadership changed, and the teams became complacent. They stopped adding unique ID elements to their fields for easy tracking, updating measurement plans, and maintaining their field tracking reports. They moved to a new tool that promised automated tracking.
It didn’t deliver as expected. All that happened was that instead of labelling unique identities for fields in their dataLayer, they had to start configuring the same identification in the tool retrospectively. None of that data or work could be shared in other tools, and they steadily became locked in. The new tool got snowballed, and they began moving to a 3rd tool without any of the updated measurement plans to support them and with the additional tasks of auditing and reworking all the patchy and now incomplete tracking. To say that it was expensive learning is an understatement.
Specialisation Without Silos
Let’s go back to our Marketing Stack. Analytics sits firmly at the foundation of any marketing campaign or web development change.
Source: Portent. (n.d.). One Trick to Set Strategic Goals
To stop circling and moving back up and down the Analytics Maturity Curve. We need to both encourage competitive specialisation and simultaneously remove siloed thinking. So, how can we do that? What’s the best compromise?
A marketing agency or internal department can be siloed in many ways: sometimes the department itself from other departments, but often as per our stack, it’s by channel or element. One of the ways we can allow for specialisation and still avoid siloed thinking is to ensure matrix teams as part of our projects.
Matrix teams are cross-functional teams that bring together individuals from different departments or areas of expertise to work on a specific project or task. They allow for collaboration and knowledge sharing while avoiding siloed thinking.
We can also create roles that focus on what we want the output to be, such as Head of Personalisation and the traditional CMO, but also define the difference between departments. This does go some way to explain some of the more lofty-sounding titles.
Analytics and Infrastructure (as per the Marketing Stack), or your web developer, depending on the company and team, are not just a department. They need touch on everything, so they are core enablers for other functions. Let’s look at how most digital agencies are organised.
Agencies offer services to Marketers; some agencies offer jack-of-all-trades consultants with varying competency levels. However, most focus and specialise in paid, social, content, or SEO. The agency may have a web development business, but usually, that’s an entirely different company. If you are a small company, it’s external. If you’re large, it’s an internal department, often run very differently from the marketing team. Then we have the Project Managers, Account Managers, and Salespeople. On top of that, some form of leadership. Analytics often comes out of either web or paid. Paid because that’s why Google bought Urchin to prove the value of AdWords through Google Analytics and Web because they’re given the job of putting stuff on the website.
How to Avoid the Divide
Source: Mcktui Consulting. Traditional Divides.
This structure creates a common divide between Web developers and Marketers. For example, to set up web analytics correctly, Marketers need help on how best to structure the data being collected by the tool, especially with the new GA4. How do we effectively transition from Event, Label, and Action to value pairs or parameters and variables? The Data Analyst will probably need help from the Web Developer to add the correct tracking. Often this work will need to be completed on a one-off basis with very little documentation or governance or additional support in a siloed structure. Our example shows Earned and Owned, like Content and SEO, sitting separately from Paid and Analytics, with Web and management functions separated again (Project Managers, Managers, or Account Managers).
Source: Mcktui Consulting – Suggested Structure.
A much better way to organise our stack is to ensure our most critical departments are structured as foundations touching all other departments, rather than as additional silos. In our Suggested Structure, we have Analytics supporting Channels alongside IT infrastructure and Web Development, with a necessary support barrier of Documentation and Governance.
This would include changes to the websites or apps and content. At the same time, they are a unifier across channels and supporting real omni-channel marketing and personalisation. They can collect good data, give information, build knowledge and wisdom, and offer a real impact. The organisation as a whole can be infused with data-led decision-making and avoid the serious disruption of infrastructure and tracking breaks.
Of course, to do this, we need to stop chasing our tails and sacrificing short-term wins for long-term gains.
Excellent, now you know who you need and how best to organise them.
You know that F**Ked data isn’t all your fault (unless it is), and you walked into a situation that was already poorly organised before you got there, but can and should work differently.
Next up:
PART THREE – WHERE DO WE GO FROM HERE?
