Sharon Flynn is a Digital Analytics Pioneer with over 15 years’ experience driving business decisions with data. Experienced presenter, data advocate, model builder from excel to SQL and RStudio. She specializes in Digital Analytics-Adobe Analytics (Omniture,) Adobe Audience Manager, Test & Target, A/B Testing Comscore, Google Analytics, SEO/SEM, and Customer segmentation.
She is the Lead Digital Analytics Consultant at Infotrust, and formerly worked as a Senior Digital Analytics Manager at BMO Financial Group in Canada.
S.F.: Making digital data easily accessible is indeed important in that process. It should be a collaborative exercise between your digital analytics team and your stakeholders. People are subject matter experts in their product or their campaign, so they should have an equal voice in the conversation of talking and thinking about it and acting on the insights. If I'm the only person in my team with the authority or skill to do that, we're going to miss out on an enormous amount of context from the person who was actually leading it. My job is to make sure that all relevant stakeholders get the data they need as fast as possible and that they can access it.
There comes a lot of responsibility with that, as well. If you are having people pull data and they have it delivered to them in the format they need by reducing the barriers, that data better be absolutely perfect.
That’s where data automation comes in. There are helpful tools out there that help reduce the human hours spent on ensuring the data is correct. They help with constantly checking that the data we thought we were collecting is what we're actually collecting so that the analytic human brain is able to interpret it correctly.
Invite data governance and data transparency in a self-serving way that motivates you. The more people who understand what you're doing, the more likely you're going to get the budget or the human time to ensure that enterprise marketing data is collected cleanly. Reducing manual work can help you save up to an entire headcount in working hours.
Setting up your data capture right makes it worthwhile investing in an implementation expert. No matter the tool or the method you feel comfortable using, document your analytics. It will greatly reduce the time spent digging around for data and it will help you make decisions based on facts, rather than gut instinct.
One weird thing I find is that analysts don't analyze their own job. We do that a lot at BMO and we are able to map all our colleagues on a digital maturity curve individually as if they were clients in an agency. Having metrics to describe the use of our metrics sounds a bit much, but it makes it easy for people who don't do what we do for a living to understand what our impact is and what we're doing. Where do your digital analysts or digital experts need to be embedded? How can they advance the whole organization?
It also helps us better understand where we are, see if users are falling back or not engaging That's where empathy comes in. Are they underperforming? Why? Are we clear enough in the objectives that they're setting? Are they empowered to set and own those objectives, or are they being dictated to them?
S.F.: It all depends on the organization’s data maturity. Typically, being more confident about marketing data in an enterprise or a smaller organization starts with having data governance and a great discipline in capturing marketing data. It’s important to gather data correctly, to monitor your KPI's and your important data correctly. Once that is in place, you should have confidence that the data is correct and not firing multiple times.
Once that process is running smoothly, ‘how to get more confident about data quality is no longer the question people are actually asking. What they're asking is how to understand how an entire ecosystem interacts to drive certain actions. What actions are you either taking, not taking, or feel ill-informed about taking right now, or what action would you take in the future?
What we need to understand is how to reconcile the difference between the currency exchange, the Rosetta Stone: what is the Delta and how can we explain the Delta. Once you do that, one of the fabulous outcomes is better understanding what your inventory of data is, where your duplications are, where you might want to retire certain approaches because you have these duplications, where you have unique datasets. Now you're really getting into efficiency and leanness, and then you can start bringing these different books of records into one place.
For example, let’s say we had 500 applications completed on the website, but only 300 were booked. The issue here is not that the 500 and 300 don't match. The issue is the 200 drop-off which you've identified as a problem in the martech technology. You know X percentage needed to come into a branch for various reasons. And now you can actually use that to answer: how do we turn the 500 instead of the 200? Why are the 200 and 500 existing? That’s the Delta, that's reconciliation. A
As subject matter experts, it's our job to interpret these questions that come from our stakeholders and actually apply our knowledge and our understanding as to what they're asking empathetically to get to the actual insights.
Not everyone needs to be a 7 out of 10 or a 10 out of 10. We have some wonderful teammates with different needs. If someone comes from Brand or Corporate, for example, they mostly need to know if their marketing campaign or social media is working solid. Data is not their core focus, they mostly need to know where the effort needs to be focused.