How do we use data to enhance our marketing capabilities? An interview with Sheila Bano Hernandez
Working within advertising, it’s difficult to separate yourself from the importance of data
But what can we do to take this further?
While so many marketers have data analysis built into their day-to-day role, being able to build dedicated teams of data analysts enhances our offering and enables specialists to do what they do best for our clients. For a growing agency or in house team, taking on this capability can take a good campaign to new heights.
To find out more about data’s role within marketing and advertising agencies, and how we can harness data to accelerate performance across all forms of campaigns, I caught up with my new colleague and data analyst, Sheila Bano Hernandez.
Where did your own passion for data come from?
My story is a bit different to your standard route into data. I actually studied business and tourism, but I've always been an advocate of efficiency and tried to do things in a smarter way, which is more or less how I started. I was doing an internship in a logistics company and they really didn't have much for me to do. I had a lot of extra time to fill and I noticed that they had some really horrendous, repetitive processes in place that had to be done every day. Because I had this extra time I started thinking;
“Is there anything that we can do differently to improve this?”
So, I began teaching myself coding and that's where things started! I automated a few things for them and I built some calculators to help them do things faster and just not have to spend hours a day on processes that could be automated.
You must have been the most valuable intern they had!
Oh, I don’t know! I’d like to think that, but they have really good people! It was really just because of the circumstances – I was new to the country and back then my English wasn’t very good so they might not have seen value in putting me to work. It was just the right time and place – I had the free time to really look into it and learn about coding (which I knew nothing of before). I had no idea how to automate anything and so I just I just found, like, my calling in that internship. Data is a universal language!
After that I landed my first data analyst role back in 2014. My only knowledge was based around VBA, the Windows application programming language, so I was looking into what was happening behind the scenes. I worked on automating loads of processes and created things like an industry calculator to compare competitors’ rates. This was great for them, but as it was very basic and just based on VBA, which isn’t widely used, it was quite rudimentary. However, it opened the doors for me to jump into my next role at an insurance company, which is where I really started to get to grips with different programming languages (SAS, SQL, Java, html), working with data flows and ultimately influencing and amending massive business processes. It was where things got real, but my starting point was so rudimentary – I built all this knowledge up just from being bored at the right moments, I guess!
This is your first position working with data in a marketing agency. Are there differences in how data can be applied in different industries?
In my previous role I was able to do a 12 month secondment in the digital marketing department, which gave me first-hand experience of working with marketing. Obviously, this was just related to the insurance industry, but it gave me insight into what data could do when combined with marketing. Best practices will always be roughly the same, as how you query the data will have few changes, but things like how complex the model is, the data flows themselves, and where this data comes from are where the biggest differences are.
Before, I knew that I could build a process and it would stay the same, be long lasting and wouldn’t need any big changes in the years to come. When working with marketing and advertising data, you don’t have that stability. With so many updates and changes in the industry, you need to always be thinking ahead and trying to make your processes flexible and adaptable to outlive these updates.
I guess that would be a benefit of working with an agency – having an outsourced team dedicated to knowing these updates?
Yes, exactly. You can’t anticipate everything but having the knowledge of the industry, the time to dedicate to your specialism, and being surrounded by colleagues who can give insights onto the changes helps to build that picture. It’s a good solution to any business who just don’t have that same resource in-house.
In an industry typically thought of as creative, how do you make the case for data to be a key driver of marketing strategy?
I really like this question - I have so many examples! Creativity is key, but you need to know if the results of your campaign are what you expect because, sometimes, you might think “this campaign’s going to do really well”, but you need the data to back it up once it's done.
In previous jobs, I've had examples where we launched different campaigns thinking our results would be massive and actually, they ended up attracting the wrong traffic. Because of this, conversion was nowhere near where we thought it would be because we hadn’t created the campaign for the kind of people, we were attracting. It wasn’t what we thought would happen, but the only way to assess this is by looking at that data, seeing why performance wasn’t as expected, and takings learnings for the next campaign. Does that answer the question?
Yes, I think it does, do you mean that by using data to build a picture of a brand, you can use your creative energy where it counts?
Yes! I once worked on a case with two brands who were marketing the exact same product so, in theory, performance in their respective campaigns would be the same. These brands were so different in terms of the audience they had and the one with a different audience to what was expected performed really poorly, which caused problems. The brand was attracting different people, and really needed its own bespoke campaign, which we identified by looking at how people were moving on their website. We identified that the customers were so wildly different from one brand to another (even within the same business), that we’d never get the same results with that product and campaign strategy.
So it helps people understand what their actual users are and maybe tell some hard truths sometimes?
Correct, even if it's completely different to what they expect. It can be a hard message to deliver sometimes, but really it’s the data which is highlighting these facts and I just have to be the messenger! It’s a good case for having data analysis involved in your process from the beginning instead of calling us in to find out why things went wrong! Even if we identify areas which will cause problems, we can look at a different way of doing thing to make it work.
How does having a team dedicated to data solutions set an agency apart from their peers? Surely specialisms such as PPC have data ingrained in their discipline?
When people think of data, they often think of it as a project to be done and not revisited but the reality is that it should be part of a process, which is where it really comes into its own. Having a specialist data team within an agency lets you build that process and apply it across campaigns to make them more efficient and accurate. Being able to have a go-to data specialist means you know someone is checking your model and making sure everything happening in your industry and marketing in general is being taken into account. If you don’t have a dedicated team to do this, you will end up with account managers and analysts trying to make sense of something outside of their specialism while not being able to do their actual job of using the insights to improve performance.
Let’s touch on data visualisation - how can smart reporting change client perception of their marketing performance and help agencies enhance their performance?
Reporting can really help change the client’s perception of campaign data and help them see and understand new opportunities. The human eye is not designed to digest data really, so we’ve always known things like tables or raw data are needed in traditional reporting, but by looking at that, we we can't possibly know what's going on, and that's when more advanced visuals come into play.
Being able to have something like a one-pager dashboard that tells the full story of what happened last week, or month by a quick glance through becomes a really powerful way of getting stakeholders on board with a campaign and reassuring them. By showing the trends visually we help boost the understanding and faith in a campaign, helping our specialists earn their clients’ trust to do bigger and better things.
With Google set to eliminate third party cookies in Chrome in 2022, most advertisers are approaching a huge data gap. How can we, as advertisers, overcome this?
We are essentially approaching using a new type of data for analysing campaigns, such as looking at how users interact with a website as opposed to information about them as an audience. Google Analytics 4, which I delivered a webinar on recently, is built around this idea of recording events with user pathways, and with the platform based on machine learning models, they can now predict what’s happening very accurately.
As an agency, we’re investing in our tech stack to include solutions such as Flashtalking so we’re able to give our clients the opportunity to own first party data which they can use instead. It’s something we’ve been doing a lot of work on and I am only one piece of this approach!
With so much of working in data surrounding automation and building processes more efficient than manpower alone, could we be looking at a completely automated future?
I wouldn’t say completely as there will always be that human element. I think we could get very close to an automated future, but people will always want to be on guard and work side by side with these processes. I think most of us would be very resistant to having robots control us, so there will always be human components to supervise what these programs and machines are doing. Machine learning is moving quickly and will change things a lot just down to the speed at which it’s evolving. There's something called unsupervised learning models which is literally the model teaching itself as it goes, but there is always going to be a component of us humans, us thinking beings, just making sure that everything's working as expected.
You’ve joined ClickThrough Marketing at a key time of growth and innovation. Do you have any ambitions of what you would like to bring to the new department?
I think it's going to be a very positive synergy. We have so many ideas to improve how we use data and free up specialists in other departments to succeed in what they do. There are so many plans, around predictive analytics in particular, as this is where there’s always so much to learn, especially in such a rapidly changing industry.
At the moment, my main focus is around automating our reporting and looking at past data to explain what’s been happening and identify how we can replicate the great results we get. The key here is, “how do we use that data to look and predict what the future is going to bring us”, and so predictive analytics is going to be the focus in the years to come. I’m very, very excited about this – showing everyone how we can use data to maximise our performance.
Interview by Megan Carthy – content and marketing executive, ClickThrough Marketing