I bought my first house many years ago, during the aftermath of the 2008 great financial crisis when my employment stability happened to be good and prices had taken a dip. The place served me well as a young bachelor in those days, but life circumstances and needs change. With a life partner, two sausage dogs and several large boxes of LEGO® bricks that are always ready for Red Tangerine’s next LEGO® Serious Play® workshop, it was high-time for an upgrade. As well as co-authoring the book, Facilitating Professional Scrum Teams, the process of buying a house came to be one of the dominant activities for me in 2023.
A quick Google search shows that anything from 25%-40% of house sales in England and Wales fall through before completion. My recent experience of trying to buy a house has nudged the stats towards the higher end, though the causes of each were quite different.
In late 2022, my partner and I had found what seemed to be the perfect place. Our offer was accepted and I soon secured a good offer for my own house. But then, the disastrous mini-budget of Liz Truss and Kwasi Kwarteng sent interest rates soaring, stopping my buyers from getting the mortgage they needed. The sale of my house fell through, and with it, the purchase of the one that we wanted.
The autumn 2022 mini-budget is an example of an assignable cause. An assignable cause is one where there is a clear explanation for the outcome, and it is usually something outside normal system behaviour. It is not usually something we can (or should) try to plan for.
Fast forward several months. I eventually receieved another acceptable offer on my house, and my partner and I soon found another place that we liked. So the formal process of buying and selling began anew.
On average, buying a house in England takes 12 weeks. However, there is going to be variations. Mortgages need to be secured, and some people may need to shop around to find the best deal for them. Or the mortgage company’s valuation of the property does not match what was offered. Solicitors conduct searches on the property, they find out about important information like flood risk, or public rights of way for example. The exact information can vary from property to property. Structural surveys should be carried out on the building itself. There is variability in the availability of surveyors, finding a time for them to conduct the survey, and then they have to produce the report. The results of either the searches or the survey could unveil new information that could cause delays, renegotiation, or even the prospective buyer walking away. It is only when contracts are exchanged that there is a true commitment from both the seller and the buyer for the transaction to take place, and this is something that happens late in the process. Until that point, there are many risks that the sale falls through. Add in that a sale and purchase can be part of a long chain of transactions, with each one dependent on the others, and the risks of delay or even collapse grows exponentially.
In a complex system like this, barring rare exceptions of assignable cause, the outcome is dependent on the variance of the different parameters within the system itself. Using the average of 12 weeks to set expectations for completion is therefore extremely risky and doesn’t tell the true story of what might actually happen. Sometimes everything goes smoothly, but oftentimes things take longer than average, sometimes much longer, but there is no single thing that determines the result. It is a combination of all the variables of the system and how “lucky” you are. This is chance cause.
System Behaviour and Human Behaviour
The more variance in the system, the more chance there is that there could be a risk of expectations not being met. Frustrations may rise, and this in turn can give rise to undesirable behaviour.
In our case, we were in the middle of a long chain of transactions, and our luck was not in. Surveys found issues. Renegotiations had to be had. Certificates were missing. Emails were sent but never arrived. In short, there were multiple delays and the process took far longer than anyone wanted. People involved started getting frustrated as their expectations were not matching the reality. And this is when the real problems started.
Different parties in the chain started to give ultimatums, threatening to pull out if completion could not be done by a certain date. These deadlines turned out to be arbitrary, our buyer moving their deadline repeatedly when it became clear that each one was not going to be met. This reminds me of how many software development projects are run, where arbitrary deadlines are often given to developers who are pressurised to meet them. When the deadlines prove to be unrealistic they are simply moved, revealing the meaningless nature of them.
Many estate agents earn their commission on the completion of a sale, so it is in their interests to get the deal closed. With this in mind, it is unsurprising to see the behaviour of the estate agents in our case as time went on. Both our seller’s estate agent and the estate agent acting for the sale of my house asked us multiple times to “put pressure” on the solicitors to make them work faster. Pressure was put on by our seller’s estate agents to cut corners and skip more detailed investigations, in essence asking us to take risks for the sake of speed. For example, when our survey on our seller’s house advised further investigations into potential damp issues and non-conformance to buildings regulations, the estate agent dismissed them as nothing to worry about and pushed us to complete rather than have them investigated properly. We insisted on further investigation which revealed that damp was a big problem and led to renegotiations on the price, but this of course all took more time.
As time went on, frustration grew and communication deteriorated into blame, bluff, threats and more ultimatums as trust completely broke down and professionalism went out of the window. Further parallels to how some software development projects are managed are there to see. With several issues still unresolved, things escalated, behaviour became more unprofessional, and we ended up breaking the chain and walking away.
I actually don’t really blame the sellers, our buyers, the solicitors or even the estate agents involved for what happened. They were all simply acting or reacting in the context of a system riddled with risk and variance that has a high failure rate. While systems remain the same, little will change, regardless of who is involved. Expecting anything else is wishful thinking.
When timelines start to slip, a common human response is to unfortunately pressurise people to “go faster” or to cut quality. This is an emotional response comprising frustration, fear and panic. It is a recipe for disaster, yet this is how humans behave if the variables of the system are not favourable and incentives are tied to speed of delivery rather than the best outcomes.
When failure rates are high, when outcomes are continually undesirable and not down to assignable cause, then the only options are to set expectations based on how the system actually behaves, and to evolve the system to be fitter for purpose by designing out causes of risk and delay. This involves systematically analysing and understanding the system, and experimenting with ways to improve it.
In other countries, the house buying system has evolved different rules to England and Wales. For example, an initial offer is legally binding in some places, and can only be reneged upon if detailed enquiries reveal issues with the property. Essentially, the act of commitment is earlier in the process resulting in a failure rate of completion that is much lower than that seen in England and Wales.
Thinking in systems, understanding commitment and different types of risk is part of “kanbanizing” a system. At its core, improving systems is what the Kanban Method is all about. The Kanban Method is so much more than visualising work and applying work in progress limits. It is a change management method that can be applied to all kinds of complex systems – not just software development. My wish for 2024 is to see more people learning about the power of Kanban.