Aller Media's Digital Analytics Manager shares his tips about marketing data quality, data foundations, campaign reporting, data layers & digital analytics
The main issue for many companies is, and always will be, the data quality. There are many reasons for it, some coming from the differences in the setups, legacy issues, but also constantly changing technological and legal landscape. A fair amount of quality issues might also come from lack of communication within the organization. I’ve seen many times that business-vital data has stopped flowing due to the changes that were made by the developers and not announced to the marketers.
I would always advise to consult all changes internally, assess their technical impact on website’s performance, and then try to set it against the true benefit for the business. It might turn out that some solutions or scripts sound nice only on paper or on vendor’s powerpoint. Another thing is to double or even triple check every change. Especially now in the Cookie Management Platform panic. Wrong implementation might be painful for the data. I’ve seen rushed installations that either stopped the tracking completely, or that were resetting the cookie at every hit - both happened regardless of user consent preferences.
Go beyond the required parameters for campaign reporting. That will help to go a level deeper in the analysis. Knowing exactly how a certain campaign was labeled will help quantify its true conversion and return on investment.
First tip is a coherent and precisely defined UTM policy. In my organization there are several units that handle campaigns. Some go on Paid Media, the others on Social Media, email platforms or the Editorial pages themselves. In order to know precisely what worked - the UTM structure has to be steel-solid.
Second tip - go beyond the required parameters. That will help to go a level deeper in terms of analysis. Knowing exactly how a certain campaign was labeled will help to quantify it’s true conversion and return on investment.
Third tip - use campaign information in the data layer. Let’s say that the UTM policy was steel-solid, yet in the reports we still see ‘(not set)’. A good way to go around it would be to populate data layer variables with the campaign-specific elements. Then, regardless of losing the UTMs along the way,we can still keep all the relevant information throughout the session or purchase flow.
Fourth tip - spread the knowledge in an easy way. Often, campaign managers are not analysts. In order to make their work easier, a good idea is to expose the data to them in a way that would allow them to act on it effectively. A nice and visual way of doing so is creating dashboards that would show only the key relevant information that they need. This way shortens lots of back-and-forth in communication and allows the stakeholders to understand the impact their actions have on the business more deeply.
Without a solid data foundation all digital activities, including marketing, are just thrown in the air. One thing is to collect the data in a coherent and standardized way. The other is to be able to evaluate the past actions and adjust the future actions. One cannot really happen without the other. Especially, when we are talking about budget optimization. If the setup is coherent across the organization, there is a clear and transparent overview of the budget utilization on a higher level. Without it - everything is siloed and cross-comparison takes additional time, resources, and often risks that the truth cannot really be found.
Without a solid data foundation all digital activities, including marketing, are just thrown in the air. One thing is to collect the data in a coherent and standardized way. The other is to be able to evaluate the past actions and adjust the future actions. One cannot really happen without the other.
Another one is the income - both from Banners and our Ecommerce. This one tells how are we doing financially. It’s important to follow it across time and brands - to be able to react in real time if certain dependencies are observed.
The third would be our cross-brand promotion. This is an important score that we focus on which tells us how good are we in cross promoting the same content category across our brands.
In my opinion, the first tactic is to listen and understand what are the problems coming from the digital area in the organization. In larger companies that can take some time, but it is important to listen to the stakeholders and find out what exactly are their challenges. In an analyst’s mind those non-technical situation descriptions can already draw a road map to technical solutions. Which brings us to another tactic - to provide MVP-style solutions. A common mistake of technicians is that after hearing the requirement they try to build a master solution that solves all the problems at once, and at the same time because of its complexity - is completely unusable. I say start simple, work closely, and develop the solutions based on constant feedback loops. In this way you can assure that your solution aims towards the defined requirements, and at the same time you “coach” your non-technical partners about the technical possibilities ahead of them.
Another thing is to keep the bigger picture in mind. In my experience, some of the problems flagged in different parts of the organization had the same high-level source. In those cases going for short term solutions doesn’t solve the root of the problem. Keeping the bigger picture in mind and working with the short term problems towards solving the long term can also expand the cross-departmental awareness of mutual dependency. Which, ideally will make people work closer together.
Preach the value of the data. Some analysts are very good at what they do, but very bad at showing how relevant their work is for their organization.
The third thing that comes back to communication - always tailor your presentation to the viewer. In my experience, even the most brilliant ideas failed when they were not understood by the decision makers. Analytics professionals have a bad tendency of speaking to the clients or non-analytical stakeholders in their own lingo. I would strongly advise to try to keep the communication as simple as possible, and to explain things that might seem obvious... Sometimes we might think that our knowledge of how things work in a backend is impressive. It is not for non-technical people. When there’s something they don’t understand - they get anxious. And when they are anxious they are less likely to buy when they are an external client, or sign-off if they come from the same organization.
Preach the value of the data. Some analysts are very good at what they do, but very bad at showing how relevant their work is for their organization. This relates to what I said previously - they tend to keep it so black boxed and mysterious that nobody really knows what they really do. I would say preach how the data is important in all the key aspects of the business. From promotion to sales, from user engagement to user retention. This preaching should be followed with specific use cases that would show how the analytics based activities drive the business value.
I’ve heard people saying they’ve boosted client’s sales by 50% by setting up best-practice analytics. I’d say, cool you’ve helped, but there were also other factors that helped the client in that case. We never question those statements, which can work as a harming factor for the truthfulness in our community.
We tend to be tool-biased. That can be observed especially with agencies that have partnership agreements with certain tool vendors. All their existence is driven by usage of specific tools, which results in the solutions that were tailored not to the actual business requirements of the clients, but usability of selected functions of a specific technology. I’ve never heard anyone commit to a mistake where due to agency’s revenue-driven approach, the actual value for the client was delivered only to the extent provided by the tool they used, and not to the full capacity that the situation allowed for.
Another thing would be that we over-value our achievements. I’ve heard people saying they’ve boosted client’s sales by 50% by setting up best-practice analytics. I’d say, cool you’ve helped, but there were also other factors that helped the client in that case. We never question those statements, which can work as a harming factor for the truthfulness in our community.
I’d say given the whole ITP discussion, more services will turn to log-in only. For analytics it changes the game in terms of how are we used to track, but also opens opportunities for more customized tracking that focuses on defined and specific users. That can result in more relevant utilization of digital resources, where the solutions will be tailored to meet the expectations of the likely-to-pay users.
The other thing will be more App+Web approach. Google’s new analytical environment will turn the analytics efforts to integrate users behavior across websites and apps - opening a whole new world of opportunities in terms of driving engagement, loyalty, and retention. This relates directly to the ITP solution, as when the companies will have to take strategic decisions in order to transfer their wider, but mostly undefined online user base to the app-centered one. This of course will not happen overnight, but I have already started to hear those conversations emerging in the community.