Data Changemakers 11: Dr Lorna Walker, Marketing and Analytics Expert
The Data Changemakers series is a set of interviews and interactions with people who have spent their careers working in or around data and data management initiatives. They have a vision for the data journey and we want to understand what they have learnt and how that drives what they do today. What are their war stories and what advice can they give others embarking on the journey?
Dr Lorna Walker is a digital marketing and analytics academic and consultant with more than 20 years’ experience. After a decade in industry working in direct and digital marketing she moved to academia and now combines work as an academic with a small portfolio of consultancy clients. She specialises in helping small and medium-sized companies set up and run effective digital marketing strategies.
Please describe a little about your own background and you ended up working with data?
I’ve had a varied career, working in both industry and academia. I did social and political sciences at university, after which I fell into a career in direct marketing. I worked in direct marketing in a number of different industries, from publishing to software to travel, including spending a few years at SPSS.
Direct marketing is all about the data and always has been, so I’ve been working with data in one form or another for my whole career.
About a decade ago I moved from industry to academia, teaching direct and digital marketing and setting up one of the UK’s first degree courses in digital marketing and analytics, which I still teach on today. I also have a small portfolio of consultancy clients.
As an academic my focus is on trying to develop graduates who not only have the soft marketing skills but are also comfortable with digital technology in general and analytics specifically. We work closely with employers in the field and they all tell us that they really struggle to get marketing graduates with the right skills.
It seems to me that all marketing these days should be data driven but traditional marketing degree courses tend to be very light on the analytics and data side of things, largely because there’s a digital skills gap amongst marketing academics just as there is amongst marketing practitioners.
Would you say that you are a business person or a technical person or something else?
I’m a hybrid. I’m part academic, part industry, part consultant.
I guess I see myself as a marketer first and foremost, but you can’t really separate out marketing from data and analytics these days, it’s all part of the same thing.
I’m more hands on with the technical stuff than most marketers I know,
What is your current role and its main responsibilities as they relate to data?
As an academic I’m working with students to help them become data literate through familiarity with analytics tools such as SPSS and SAS, and helping them understand the possibilities that data unlocks for marketers.
The level of statistical literacy that students arrive with tends to be very low, so we’re taking them through the process right from the basics of exploratory analysis, significance testing and so on through to quite advanced predictive analytics.
I work with quite a few students on research projects where they’re working with company data or generating their own data sets, perhaps via survey research or through scraping the internet for data, so there’s a lot of variety in that in terms of the kinds of techniques they’ll be using.
What has been the most challenging data-related project you have worked on and why? What was your role in it and was the project a success and why?
For my PhD research I worked on a dataset of 150,000 tweets sent by MPs in the run up to the 2015 general election, using SPSS Modeler to build predictive CHAID models to determine the likelihood of politicians’ tweets getting retweeted as well as to identify the factors that most contributed to retweeting.
It was challenging because of the ‘messiness’ of the data, even though it was essentially one relatively straightforward data set it still took an incredible amount of time to clean the data up and get it ready for analysis.
Other challenges were determining what questions to ask of the data, using modelling techniques and software that I had not used before, being confident that I’d interpreted my results correctly, combining analysis of both structured and unstructured data and so on. Ultimately it was successful because I was awarded a PhD on the basis of this research.
What did you learn from the experience?
That analytics is a messy business. You rarely get a nice clean data set with everything ready for analysis, along with a clear set of obvious research questions you can focus on. It’s much more ambiguous than that in real life. There’s no ‘right answer’. The answer you get depends greatly on the way you handle the data, decisions you make while you’re cleaning it and analysing it, and another researcher could make completely different decisions and come up with a different answer. It takes a while to be comfortable with that level of ambiguity.
In a business context I think the key implication of this is that data alone isn’t the answer – it needs to be combined with a deep level of business knowledge and understanding from whoever is doing the analysis, in order to be able to interpret the findings in a sensible fashion.
In terms of the findings of the research itself, the findings were pretty depressing. I did a content and sentiment analysis of a sample of the tweets to see how these factors influenced retweeting. The idea was to be able to generate a set of practical recommendations that would help politicians gain more traction on social media through retweeting. This was all pre-Trump and political use of Twitter was less widespread than it is now.
Sadly, although perhaps not surprisingly, my research showed that negative tweets were much more likely to get retweeted than positive tweets. In particular, tweets attacking an opponent or including a message designed to stimulate fear or anger in the reader, got the most traction.
What do you think are the key trends in data management today and how do you think it will change the way we all do business?
On the marketing side I think there’s much more recognition that data and analytics isn’t an optional extra but really a critical part of doing marketing ‘properly’. In particular marketers have access to much more unstructured data in a digital format and much more able to extract meaningful insights from it.
Also from a marketing perspective GDPR is having a massive effect, with organisations really having to think about why they’re collecting the data that they collect and whether they’re actually getting value from it or would be better off not having it at all.
Consumers are more aware of the value of their data, particularly in the light of the Facebook and Cambridge Analytica scandal, and that’s going to have a knock on effect for all kinds of organisations as consumers start to ask more questions about why particular bits of personal data are needed and what they’re going to be used for.
What advice would you give to someone embarking on a large data-related project today?
I’m talking here about the kinds of projects that academics or marketers might be conducting themselves, rather than massive big data type projects.
As an individual researcher you really can’t spend too much time on cleaning the data.
Taking time to clean the data and really understand what all the fields are and how they might be useful is invaluable. It’s very tempting to get stuck in to running tests as soon as you get the data but without proper cleaning first you’re going to start generating findings that are meaningless, misleading or impossible to interpret.
It’s also critical to document what you’re doing as you do it.
This is a lesson that I learned the hard way when I was doing my PhD, as I ended up spending a lot of time repeating things that I’d already done but not properly documented, or building models that I returned to a few months later and struggled to understand.
On a more positive note, individual marketers and researchers shouldn’t be afraid to get hands on with the data. It’s tempting for marketers in particular to outsource this stuff, but no one understands your data as well as you do.
Are there any particular skills or qualifications you consider to be vital to your success?
For my personal success, not really. I think my career shows that it’s possible to have a successful career in data and analytics without having formal qualifications in the field. I don’t have a statistics degree, and although my academic work has included a lot of analytics, I’m definitely not a statistician. I’m a practitioner first and foremost, the data is a means to an end rather than the end in itself. I had some formal statistics training years back when I did an MBA, and some more as part of my PhD research but beyond that I’m largely self-taught.
If people are thinking of entering this field I think they need to be comfortable with the statistics and the analytics tools, but it’s perfectly possible to learn much of this stuff on the job. However, if you’re planning a career as a data scientist then formal statistics training is going to be helpful.
For most marketers it’s possible to get up to speed with 90% of what you need to know with a combination of natural curiosity, training courses, learning online, and a willingness to get stuck in and give it a go. Don’t be afraid of the data or of statistics.
What 3 words would you choose to describe yourself?
Curious, reliable, enthusiastic.
What are you best known for or what do you like doing outside of your working life?