The initial implications of Covid-19
Ammonite hosted our 1st ever Zoom Leaders in Data (LiD) ‘round table’ discussion. We definitely missed the incredible pancakes and bottomless cups of coffee from Riding House Café but the chat was good as always. The focus was predictable, ‘The implications of Covid-19’.
Attendees:
Host – Keith Robinson – Director, AmmoniteData
Host – Phil Marks – Director, AmmoniteData.
Chair – Janet Bastiman – Chief Science Officer, Ammonite & various.
Chris Bowen – Head of Business Data & Analytics, O2
Mark Jones – Senior Product Manager, Birdie.
Rob Barham – Head of Data Science, Engineering & BI, TheVeryGroup.
Matt Doltis – Head of Data, yulife.
Premal Desai – Head of Data & AI, TheGymGroup.
Matt Whiteley – Head of Analytics, Mobkoi.
As this discussion unfolded it became very apparent how differently the Covid-19 crisis is affecting the diverse industries represented round our virtual table and those differences are driving their individual business and data strategies and tactics. Business models ranged from traditional corporates, to those involved in frontline crisis services and from e-commerce to a pure bricks and mortar offering.
Whichever model, everyone in data is busy supporting huge changes as C-Suite struggle to rapidly refocus on shorter term revenue protection strategies by diverting, reducing or, in a few cases, increasing resources.
We are all exploring new territory. None of us has ever experienced anything like this. Sometimes it’s frustrating from both a work and personal perspective. Some issues below are familiar, but all are exasperated by the limitations of working from home and being apart from all colleagues, not just the ones in other offices and teams. Of course, the initial challenge was setting up a home working environment, but I’ll leave that for others to rake over.
Communication
Weaving its way through all the discussions was the biggest data challenge of all … communication. Things you would normally pick up on in the office, you just don’t know about. As a result, some data leaders are over-communicating to find out what people are up to, spending much of their time checking (and nagging) to see if people across the company have everything they need and that they are communicating in the right way.
On the flip side there is a lack of communication. Skunkworks are happening where the data team have zero visibility until something is deployed or teams have misused a piece of analysis that has been given to them. A great example; a team taking a daily export of a dashboard and building up their own dataset, not understanding it was already based on pre-aggregated values.
As a consequence, many of the leaders are much more proactive in their communications. A good tip – send short, instructive videos round the company e.g. a 2 min whiteboarding session to explain analytics 101 (e.g. building a data table). Also send constant reminders that the data team are around, and people should check in first when working on data-related projects.
(Infra)Structure
However, as requests to check things come out of the woodwork some early triage will be required. Now that people are at home thinking about what they want, reacting to client requests and their own creative thinking, they are throwing lots of new ideas at the data and engineering teams, including automation requests. That can be overwhelming at the best of times but even more difficult and frustrating to combine working flat out and with non-work activities like home schooling, walking the dog and eating!
It’s also been challenging to ensure people have access to the right data to do their job and use the approved technology. In one of the more traditional businesses it emerged that several parts of the business have their own shadow ‘data infrastructure’ – an Access databases and VBA code running on Excel. Others have rogue purchases of other technology. Some of this is now breaking down as people don’t have access to the same data and storage as before, like a desktop at the office they were running it off. When things calm down, there will be an opportunity to have conversations about rebuilding in a more sustainable way. But at the same time, there is also a lack of appreciation from C-suite of the risk of the large technical debt being built up.
There are also requirements for rapid development and creativity. In one front-line carer platform business, they’ve had to quickly develop robustness and new analysis to support carer decisions. With very traditional, paper-based customers they are now trying to digitise something that has been lagging behind for some time.
Industry
Of course, many businesses and industries have been devastated. None more so than those with physical service offerings like restaurants and gyms. Going from market leader to zero overnight, in full on crisis mode, focuses the mind. There’s a massive shift from a longer-term time horizon to short term tactical efforts. Figuring out how to deal with not being able to provide a service to your customers anymore, with a vastly reduced workforce is a unique problem. It’s also very hard to forward plan when you don’t know how long this situation will go on for.
However, some industries are incredibly busy. O2 has seen a 50% increase in voice traffic in a 2-week period. They’ve also been working with government to provide anonymised data to check if people are abiding by the new stay at home rules. It will be interesting to analyse other new customer behaviours in the data, but from a comparative point of view, data scientists will probably have to ignore this current period which is causing a huge data blip.
In retail 75%-80% of revenue is spent in stores so it’s easy to imagine a large drop in consumer spending over the coming weeks. But there is still a large benefit to online and e-commerce businesses currently. A lack of other disposable spending opportunities might also provide a short-term boost. In a future world will this demand subside or continue? It might be quite hard to go back to old behaviours, which will have all sorts of operational and organisational implications.
1/3rd of attendees are using and recommending OKR’s (objectives and key results). It has “lessened the impact of remote working and increased transparency because we know where to focus our efforts at team level”. It’s also a time for project management disciplines to shine.
Weird, busy, frustrating and scary times ahead. Stay well, home and safe. Love to the NHS.