Great show & tell from Welsh Data Science Graduate programme
Last week I had the privilege of hearing the stories of Welsh Data Science Graduate Programme members. These were tales of their first industrial placements.
I’ve shared previously the news that Laughlin Consultancy is collaborating with the University of South Wales to support their new Welsh Data Science Graduate programme. That gave me the opportunity to attend this very encouraging ‘show and tell‘ time.
Each of 13 students presented to an audience of their peers, employers & academics. We were treated to brief summaries of what they have achieved on their first industrial placement. It was truly impressive to hear what they have done in only a few months.
Discovering the reality of how businesses use Data Scientists
As I listened to these enthusiastic & engaging young people, I realised what I was hearing would also be of interest to you, my readers. I say that because I know a number of leaders who want to know how others are using data scientists.
So, I started listening for themes. How were the Welsh employers (given the opportunity to have one of these student Data Scientists) using their skills? In this post, I will share with you the common themes that struck and impressed me.
Software being used by Data Scientists in normal businesses
That slightly provocative heading is to distinguish these placements from high tech companies or those who impress on LinkedIn. A good range of leading Welsh businesses were represented, mostly in Financial Services, but a few others too. Some were obviously more mature in their use of analytics, but some were starting from a low base.
Each student shared the software they worked with for their placement. So, I started keeping a count. For your interest, here are the top software cited:
- R (almost every student used this, even if they were the only one at their business)
- SQL (almost as popular and for a number of students a new skill for them to learn)
- Python (only about half of the students)
- SAS (and amusingly a few used it mainly for PROC SQL – which is hilarious overkill compared to native SQL)
- PowerBI & Tableau (much less often, but a few for Data Viz)
Numerous packages were also mentioned, mainly the R and Python packages I’ve recommended before. No mention of using Julia yet.
Developing their business communication skills
As someone who trains successful analytics teams in a range of Softer Skills, I was reassured to hear an emphasis on communication. As well as demonstrating their comms skills in their presentations, most cited opportunities to present.
In one Q&A their focus on developing “business acumen” skills was also cited. By this, they meant their ability to understand and use language relevant to their commercial placement. Almost every student mentioned visiting other parts of the business to interact with stakeholders. Such an important skill for their ongoing stakeholder management.
Most of their presentations also showed greater creativity than your average boring business slideshow. 3 of the students recorded a video instead of using slides – which really helped us all avoid “death by Powerpoint“.
Their challenge to keep their updates to (roughly) 5 minutes was also important training for real-world analytics. I’ve mentioned before the work of The Brief Lab & I think Joe McCormack would have been impressed with their brevity.
Getting their hands dirty with real data
It was also good to hear a number of the students mention their learning from working with messy data. I’ve posted before on the challenge of how much time can be taken up with Data Wrangling.
So, I was pleased to hear that mentioned, as well as placements that worked on Data Quality & Data Dictionaries. Hopefully, this will serve them well on new projects as they value time avoiding GIGO.
Tasks working on cleaning data & managing data quality, as well as maintaining metadata were all mentioned. Vital foundations for robust modelling work.
Improving processes and dashboards, incremental change
Another sign of the pragmatism and reality of these placement stories was their focus on evolution rather than revolution. I’ve worried out loud before about how many businesses were hiring data scientists without knowing what to do with them.
It sounded like these students were helpfully focussed on pragmatic improvements. The most popular applications of their skills were improving existing dashboards or business processes.
Realising the need to make gradual incremental improvements that make the financial case for more significant investment/change is useful learning. Hopefully, students will continue to have the opportunity to learn where their analytics work can have “biggest bang for your buck“.
Technical work focussed on Propensity Models and Segmentations
Although there was a diversity of technical analytical work shared, two techniques were repeated themes. Those were customer segmentation and propensity models (most often using logistic regression).
So, out of all of the potential trendy algorithms & innovative machine learning methods, businesses found well-established stats the most useful. That echoes my experience with clients. Data Science students do need to be given an opportunity to learn traditional marketing analytics stats methods, not just new variants.
Helping with culture change, innovation & evangelism
Given the enthusiasm and eloquence of these trainee data scientists, I was not surprised to hear that they were enlisted to help with culture change. A number of the students shared that one of the reasons they were out talking with stakeholders was to ‘spread the word‘.
As bloggers have mentioned a few times on this blog, culture change is one of the key challenges for Analytics & Data Science leaders. I’m pleased to see these students ‘thrown in the deep end‘ with encouraging data-driven changes & innovation.
Hopefully, those students might find some of the lessons shared in my recent series of audio interviews with Data Science leaders useful. I recommend they listen to:
So, how could you use Data Science graduates?
I hope that insight into the good progress from the University of South Wales Data Science students was interesting. I am certainly encouraged in my collaboration with Welsh Data Science Graduate programme.
What about you and the talent pipeline for your business? Are you building relationships with local universities? Collaborating in placements like this could really help you learn how you might grow & apply Data Science capability in your business.
For any leaders in South Wales, I encourage you to get involved. By working together we can help ensure that the next generation of Data Science students have the skills they need to make a difference in businesses. That would be a great result for our society & economy.