Ibrahim Elawadi is a Data Advocate, Public Speaker, and Global Director of Digital Analytics at Philips. He has a master’s degree in Marketing, bachelor’s in Electronics Engineering, and currently studying Astronomy. In his free time, he loves taking pictures of the stars.
I.E.: That's a good question. Here are some of the main reasons I believe that happens:
I.E.: Analytics should always answer a clear business objective that helps us understand the exact contribution of the analytics team to the bigger picture.
This could be e.g., increasing sales or maximizing the ROI on ad spend. You can also extend that beyond digital marketing. If you are e.g., a manufacturing organization, you could think about how to increase the efficiency of the supply chain or how to have a better prediction of demand.
Many times, there is no clear business objective in mind and this is where many challenges start surfacing. In that case, analytics becomes the target, not the medium. And that process doesn't always get satisfactory results because you end up with insights that are not really actionable.
I.E.: The clarity of the question is crucial, as well. If you take, for instance, audience research: before you start any type of research, you need to start with a clear research question. What do you want to answer?
It is difficult to take action on the data if the research question is not really clear. So, if you think about it as a scientific method, the same thing should apply to analytics. You need a clear question before you start an analytics project. You don't start looking at the data waiting for something interesting to happen. This could happen if you are lucky, but this should not be the way to do it.
A method I use for that: I start by having a conversation with the business owners to understand what they are doing and what they aim to achieve. This helps me better understand the business context better and gets closer to the questions they want to ask. If the questions are unclear at the beginning, that’s when the real work begins. It's not only the business owner or the analyst’s responsibility: both need to work together to clarify the questions and their purpose before the project starts.
I.E.: With every analytics project, you need to have a clear process for measuring its impact. The impact of analytics can be measured in different ways:
The output of analytics can be many different things ranging from the dashboards that we build to the number of reports we are producing. It can also be quantified as the number of questions that the analytics team answers with a ticketing system.
A different approach is measuring the outcome.
For example, how many people use the dashboards the analytics team is making and how are they activated? How many people in the organization open and use the reports? Are they used in the business or do they resonate with any decision makers?
The approach used is in reality dictated by the maturity of your organization. Data maturity differs significantly from one organization to the other. How we measure analytics impact depends on the size of the organization, the investment in analytics, and how high analytics is in the hierarchy of the organization.
In some cases, I would suggest measuring both. If the outcome is difficult to measure in a young organization where the analytics team is not mature yet or there are cultural challenges, a good method would be to start by measuring your output and gradually transitioning to measuring the outcome, as well.
I.E.: Another important factor is the data culture of the organization. Some studies show that roughly 92% of people surveyed say that culture is the biggest obstacle to becoming a data-driven organization.
When you think about culture, it's both about people and processes. And it's often the case that you might have a really good analyst that knows what they're doing. But there is no clear process about when they should be involved in a specific project.
E.g., if you're doing a marketing campaign, when should the analyst be in the campaign cycle and where are those people sitting in the organization? Are they sitting at the table as decision makers or do they hold the role of a specialist who answers multiple questions for projects that cater to different teams and stakeholders?
I.E.: Another important factor is how senior the data people are in the organization. And, by extension, how high are they in the organizational structure? In some companies, analysts are part of the C-suite as e.g. data officers with a strong data background. This ensures that data is treated like a first-class citizen in the organization.
Data is no longer something we consider if we have the chance, the time, or the luxury. There is a data ambassador in the organization who has a voice that is very well heard all the way up.
I.E.: That’s completely true, data without activation is just a cost. It's just an additional cost for the organization, right? A question that I always like to ask is: if you don't have analysts in the organization, what are you losing?
If you can answer that question and name the things you are losing, then those are the things that are really important and we should double down on.
I.E.: When we think about the ultimate goal of analytics, it could be boiled down to the following:
I believe that these are the three key areas that we should be focusing on. In turn, translate the impact of analytics into one of these buckets.
Does marketing data, for example, help decision makers know which channels they should spend their budget on? This is a very clear outcome that I can materialize. Does it help us build personas for our marketing team? If the answer is yes, then this is also a clear target we can work on. Are we doing A/B testing on social media or landing pages to find out which design works better? Then this is an optimization exercise and the outcome is also very clear.
So, having that in place before we start any analytics project really helps prove the value of analytics.
I.E.: A model like the data mesh can make it easier to act on insights and help prove the value of analytics. Instead of having one centralized data team that is maybe isolated from the rest of the organization, bring them into the actual team that makes decisions.
In practice, this model means that the marketing team, for example, has dedicated data people. Having an analyst on the marketing team sitting at the table when a campaign is designed, during implementation, and when doing performance analysis, helps them have a stronger voice in the organization.
Of course, this model doesn't work with all organizations. Not every organization will have the resources to bring one data person or data team to every department, but bridging this gap to the analytics consumers should definitely be a priority for analysts.
I.E.: What I also believe is really important for the analysts is to be able to speak the language of the business owners using their own language. This means using a minimal amount of technical words.
Personally, I am a big fan of simplifying analytics and the terminologies that we use. When you design your own presentation for specific stakeholders, you should use simple language that they can understand without much effort.