496,000 Go Ahead train passengers face delay or cancelled trains every week - a more significant headline

There are few London commuter's who would seek to defend Southern Rail, reputedly one of the worst performing rail operators, but as a data-specialist with a finance background I read the recent news headlines with some despair. "Southern Rail owners reveal £100m profits after months of cancellations" from The Times, being typical of the coverage last week when Go Ahead announced their annual figures.

So being curious to understand the reality, I dug into the numbers a bit, and wanted to share my findings. Go Ahead is a group, and includes includes not just rail franchises, but also bus operations, across the South East and including 24% of London routes. Bus revenue is 26% of Go Ahead's total revenue, but generates 64% of operating profit. By comparison Rail revenue is 74% of the Group revenue, but only contributes 36% of profit. So immediately the £100m headline grabbing figure is off the mark in relevance to the 'months of cancellations' for rail passengers.

The rail element of Go Ahead's business is a 65% share of Govia, which runs three franchises Govia Thameslink Railway [GTR], Southeastern and London Midland.  GTR itself is an accumulation of Southern, ThamesLink, Gatwick Express and Great Northern. Clearly Go Aheads rail interests are extensive, which made me consider whether the profit from Southern Rail was indeed excessive, in comparison to other franchises.

A useful reference for this is a report published by the Office of Rail and Road [ORR] in March this year : Passenger Rail: trends and comparisons (link below). This provides accumulated data across the UK passenger rail network, of revenues and costs across the franchisees. The report is based on data to 2014 - so is not 100% current, but provides history back to 2001 so trends are pretty clear, and the data is more consistent and complete across operators.

First highlight is the overall level of profit: collectively the franchised operators have revenue of £9.4bn, and costs of £9.2bn - a rather thin margin of £200m or 2%.

The report provides a useful chart that shows total revenue and total costs by franchise. The report doesn't provide underlying data, so an element of judgement was required to convert this into a chart that shows the actual profit per franchise; as a £m figure and % of revenue. I ranked this by the size of profit in £m so it shows those generating most profit on the left, reducing those that were loss making on the right. All data is 2014 figures.

This shows that absolute profit ranged from ~£33m (Northern : NOR) down to a loss of £3m (Greater Anglia: GRA). The percentage profit fluctuates, but is broadly in line with absolute profit [the % trend slopes down from left to right] with a few notable exceptions higher than the average: Mersey Rail: MER (14%), C2C (9%), Transpennine Express : TPE (8%), London Overground: LRL (7%). And Southern Rail? Well it's profits (remember for 2014) are in 2nd place at £30m [SOU], but at a margin of 4%, based on a revenue of ~£746m.

The report goes into considerable detail of the factors behind the components that comprise each operators revenue and costs, and it succinctly highlights the complexity of running the UK railways, and also the difficulties in useful comparative measures.

However this doesn't help the frustrated commuter at Waterloo, so I did one additional piece of analysis. What frustrates most passengers is not profit, but the combination of prices and punctuality. Most commuters feel fare rises have been relentless without measurable service improvement. So I had a look at the current data provided by the ORR for current performance.  Taking the latest data from the ORR (year ended 31 March 2016) I've calculated the average passengers per train, for each operator and the average number of trains cancelled or severely delayed each week. From this I derived the number of passengers impacted, and for reference indicated this as a % travelling.

This data represents the operators in a slightly different way (and reflects the changing landscape of franchise operators) but, irrespective of this, the chart is not good for Go Ahead. Govia Thameslink ranks highest, with approx 332,000 passengers impacted by cancelled or severely delayed trains every week. Highest in number and third highest as a percentage (5.3%). Add in the other Go Ahead franchises: South Eastern and London Midland and the total impact is 496, 000 passengers every week.

I'd been tempted to leave it there, but curiosity got the better of me. What was the trend for GTR on the weekly passenger numbers impacted by cancellations and severe delays? The result is below, and it doesn't require much of an explanation, or headline.

Sources:
Go-Ahead Group 2016 Results: http://www.go-ahead.com/en/investors.html
Office of Rail and Road: Passenger trends report: http://orr.gov.uk/publications/reports/passenger-rail-trends-and-comparisons-for-franchised-operators

If you're sitting comfortably, it's time for storytelling with data.

Like many contemporary concepts the world of analytics doesn’t stand still. Organisations are on a constant quest to do more with their data, to get more value, greater insight and understanding and do all this at lowest cost, but incorporating innovative technology.

One concept that gained much traction during 2015, and looks set to really peak in 2016, is in storytelling with data. So, if you’re sitting comfortably, let me explain.

One of the most common challenges for organisations is the difficulty in converting the insight derived from data analytics into something actionable. This encompasses the clear identification and explanation of the ‘so-what?’ element of a piece of data analysis. However brilliant and clever the analytical techniques may be, it is essential to clearly communicate the outcomes to business leaders, so they understand why the findings are of importance, to allow validation of the recommended action, and to ensure the analysis leads to a definitive business decision that impacts the business: typically via decisions that touch individual customers, suppliers, employees etc.

Data storytelling is a technique that is most beneficial when applied to convey what are often complex findings, derived from a multi-step piece of analytics. With a multi-step approach we can take business people on a journey, simplifying complexity, in a way that aligns with their emotional and intellectual awareness and that explains, educates and convinces.

Data storytelling is somewhat different to visualisation and in particular Infographics. Though the two themselves are quite distinct, as this article highlights.

Infographics have become a hugely common and popular approach to summarising statistics and can be found in all kinds of avenues, not just in businesses but also in news and media. There are some good explanations around of why infographics are so popular, and so useful at conveying information. This article does the job particularly well.

A key element of data storytelling is often visual, but it’s more about providing a guided path through findings to show how an analyst has taken some logical steps to arrive at a final result or set of options or outcomes.

It’s not surprising that many software vendors are seizing the momentum around data storytelling. Tableau have added a feature called ‘storypoints’ and Qlik allow a guided story via ‘pathways’.
There is also plenty of quality educational material to encourage good, if not best practice. Tom Davenport’s article in the Harvard Business Review, for example, is an excellent summary of the 10 kinds of stories to tell with data. And a good article in Computer World that identifies the trends in storytelling for 2016.

What I haven’t done here is to delve into detailed illustrations of storytelling in practice; again there’s lots of examples out there. Here are four that highlight a range of approaches:
The FT.com: What’s at stake at the Paris Climate Change Conference.
How far can you travel when your petrol / gas warning light comes on.
Gun Deaths in America: making sense of the numbers.
How sunspots impact global weather.

These examples provide an interesting range of examples and approaches that should provide a clearer guide to the art of data storytelling, but if you want to know more, there is a compilation of the best resources, including links to some excellent blogs.

And if you’re not convinced by the power of storytelling; do you need reminding what time Cinderella had to leave the ball? Or what Jack swapped for the magic beans on his way to the market? Or what animal made Dick Whittington's fortune? Just make sure your data storytelling enlightens and enchants and doesn't make your audience fall asleep!