Portfolio Links

Examples of my previous work, and ongoing projects live in a variety of places.

LinkedIn Featured

My LinkedIn Featured is the best place to start. It has the best samples of the varied work I do. From in-depth SQL data analysis to hard hitting data journalism that was featured on the BBC news at six.

Data Analysis

SQL

Python

  • I have quite a few Python data analysis projects I’ve done during my Python upskilling journey.
  • I use Python scripts occasionally during my day to day data analysis. More on that later.

DAX/Power Query (Or Power BI for the layman)

All of LandTech’s Regional Market Reports were built using an ETL process in Power Query. This combined 20+ datasets from public source datasets to LandTech’s proprietary planning applications data resource.

For an explanation of the commercial impact of LandTech’s Regional Market Reports (105% Net Recurring Revenue of £12 million ARR despite entrant of a fast-iterating competitor with low technical debt), read my summary in my “Selected Portfolio” here and here.

Tableau

I’m currently recreating my London Affordability story as a Tableau dashboard. I have a Tableau certificate here.

I didn’t use Tableau in my Professional Data Journalism at LandTech as its support for geospatial datasets was quite weak when I started. Continual Local Authority boundary changes meant I needed the ability to tweak and upload new GeoJSONs on an annual basis.

After a month’s trial of using Tableau, I abandoned it in favour of using a mix of Excel, Carto, Datawrapper, and Flourish.

Other

A Data Journalist is a Data Analyst that specialises in storytelling.

Every data story I undertake has a lot of under-the-hood data analysis. Here’s one example to illustrate what can go behind a data story:

LandTech’s New-Build Premium (30 million+ records, SQL, Python, Flourish)

LandTech’s New-Build Premium Data Story (2020, 2021), highlights the premium new builds sell compared to non new builds in England and Wales. It highlighted the flaws in the government Land Registry’s methodology, which compared median new builds to median non new builds. This placed the premium at around 33%.

By comparing on a price-per-square-foot basis I was able to show the premium was much lower, at 6.6%, and was declining due to the anticipated phase-out of Help-To-Buy.

This was a mammoth data analysis project. To get the price per square foot data I had to combine the house prices from Land Registry’s Price Paid dataset with the property square footage from the Energy Performance Certificate dataset. Neither of these datasets used the Unique Property Reference Number (UPRN), so combining them was a challenge, and I didn’t have access to LandTech’s proprietary geocoder.

I wrote a script in Python to create a primary key and a foreign key from the first line and post code of each address. Initial efforts provided a 60% match of addresses, due to my use of Regular Expressions not accounting for addresses with house names, and not door numbers. The script was tweaked until a 97% match was obtained. This was for over 30 million records in the EPC dataset – the entire address database of England and Wales.

The datasets were then combined in SQL, and subsequent analysis was carried out in Excel. The choropleths (or heatmaps) were visualised in Flourish.

Personal Data Journalism

Data Journalism is the art of telling stories with data. I use cutting edge data analysis tools and techniques (ETL with SQL, DAX, and Power Query) to bring out the stories which I then write about and visualise with Flourish, a visualisation tool used by newsdesks from Sky News to the FT.

If you want to read about some of my personal data journalism here (including how the price of the humble baked bean has skyrocketed), check out my blog.

Professional Data Journalism

As a Data Journalist at LandTech I used my intrapreneurial skills to pioneer the art of using Data Journalism in B2B SaaS Content Marketing and Thought Leadership.

  • LandTech’s Housing Delivery Test provided over £6 million in inbound opportunities in three years, changed the workflows of hundreds of property developers, and generated LandTech press exposure in Property Week and The Times.
  • Data Journalism and Thought Leadership on unlocking the green belt to build housing in 2021 got LandTech featured on the BBC News at Six.
  • LandTech’s EPC data story was featured in 100+ press publications, including the Evening Standard, The Independent, and LBC.

My author profile on LandTech’s blog showcases stories I did for them, and LandTech’s Report Hub, showcases LandTech’s Regional Market Reports – a series of white papers I helped to create using Power Query and Flourish. Used by Account Managers during license renewals, they helped LandTech hit their annual OKR of 105% Net Recurring Revenue.

Github

My Github is where my personal learning projects live. You can find exercises and projects I’ve created while learning D3.JS, SQL, and Python.