At the recent JAFAC conference, Ben Gracewood told the story of how Point of Sale (POS) developer Vend transformed their development organisation following a funding crunch which meant they needed to make some significant layoffs. He explained the filtering that was happening whereby the organisation was ignoring signals coming from outside, and explained the impact of the crash on the people and the organisation. He explored a number of silver linings which enabled the organisation to weather the changes and recover, and challenged the audience to consider how their organisations would cope if they were suddenly faced with needing to run the business on half of their current revenues.
Gracewood emphasized that there was a massive impact on the people, and that the organisation had no choice but to make layoffs which had real impact on people's lives and livelihoods. He started by describing how the cerebellum is a component of the brain which acts as a filter, blocking out inputs which it considers irrelevant or which could cause confusion. He stated that Vend had an active corporate cerebellum that resulted in a number of blind spots.
He called out four factors which were preventing them from seeing the reality of the situation they were in:
- Limited by team structure - they had adopted the Spotify team structure, without carefully considering if the structure was right for their specific context; the reality was the squads were working in silos with very little collaboration and sharing of information.
- They were fighting on a very broad front - by trying to build a product which could be used by any POS customer, they were unable to prioritise feature development - he mentioned having to trade off "bar tabs and table layouts against serial numbers and gift cards". They shifted to focusing on multi-outlet, inventory-based retailers which makes prioritisation of feature development achievable.
- They suffered from "MVP-itis", which resulted in under-engineered features being released because "it's an MVP"; many features were released which were not scaleable or robust, or had poor user experience.
- They were fooled by what he called the "fallacy of QA" - they had manual testing which they believed improved the quality of the product. The reality is that the manual QA approach slowed down the release cycle and had no impact on product quality.
He explained some silver linings that resulted from the changes:
- Increased talent density; with fewer people they immediately achieved more. He said that the day after people were let go they started getting more done. This was because of the talent density and the removal of bureaucracy, less unnecessary debate and less blockers. He referenced the Netflix Culture slide deck and showed a graph looking at the impact of organisational compexity on high performing people, and how we have a tendency to add bureaucracy to try to prevent chaos, whereas the opposite is generally what is needed.
- With the reduction in the number of people, the teams became more cross-functional and collaborative - there is more line of sight between teams and stakeholders and opportunity for confusion.
Gracewood also identified what he called saving graces - the things that they were already doing which enabled the company to actually survive the disruption and be able to adapt:
- They had continuous delivery infrastructure processes already in place and were able to continue releasing software at a fast cadence
- They had converted the product to a microservices architecture, which meant that the smaller groups didn't need to try and understand a monolith, they could focus on working on the smaller microservices and respond quickly to implement the needed changes
He summarised by showing some figures which indicated the improvements the organisation has achieved since the restructure in performance, support costs, help desk calls, customer satisfaction and profitability. He explained how the key metrics are made visible internally and how the teams use them to monitor and improve their own performance.
He ended by challenging the audience to ask themselves:
What would you do if you were told you had half the money you currently have to run your team or company, and why aren't you already doing that?
In the Q&A after the talk he was asked about the impact this had on the culture of the organisation, and he made the point that they already had a good collaborative culture prior to the disruption and that they worked on retaining that culture despite the losses. Management was acutely aware that the reduction had a massive impact on the people who left and on those who remained, and they genuinely cared. The reduction was unavoidable if the organisation was to survive and it was undertaken with concern and empathy.