The Harris family emerged from a pool of tears as the Homefinance app was being developed. In the early development of Homefinance, I used my own accounts to test what we were building. This worked for a while, but then some serious cracks appeared in this approach. First there were a lot of entries in my personal accounts, so it could take a lot of time to enter them; second the accounts were messy, as real world things are, so it was difficult to work out whether the strange results Homefinance was giving at times were because the accounts were messy or the code needed editing; third, when we were testing the code, we wanted to know what the expected result would be – this was not easy when using real world accounts and fourth, and I think the most important reason, was that we wanted to safely play with the data, so that we could see what happened when things did go astray, as they do in the real world.
At first the Harris family was just called a “reference dataset” descriptive of the numbers we were testing, but not very friendly. This worked for a while, but we found that as we worked on the “reference dataset”, we needed to describe what the “reference dataset” was being set up for in more human terms: what made up a realistic income flow; what types of things should be included in the expenses; what types of expenses were there, and how best to categorise them; how would a budget work; what would you do with savings?
All these questions became easier as we started to talk about them as if we were talking about a typical family who are using Homefinance. Most families have two income flows, one from each adult partner. To make it more interesting, let’s make one have a part-time salary, and the other take drawings from a family business. That provides a clear difference about how the income flows are treated, and how they might be used in Homefinance. Why is there a part-time salary? That partner is studying -which adds another layer to the support she might have as a result. The family has children, and they are learning to manage money, so new accounts need to be set up for them: the older child, Emily is using the income she gets from her after-school job to save for a car when she is seventeen; and Jack gets some regular pocket money which goes into an account for when he wants to buy a computer game, or buy snacks at his football game on Saturdays.
The same applies to expenses. The reference dataset came alive when it became the Harris Family. Suddenly there was a set of real-life expenses that a family is likely to have. Some are routine, like Groceries, or the mortgage, but then the family expanded – there were two children, one in high school, and the other in primary school – expenses now include orthodontics, music lessons, sports, school fees, birthday presents. And of course, most families have animals to share their lives with – so there are now vet costs, and pet food to consider.
Sarah is studying so she has expenses to cover as a result – so we add textbooks, travel to university. She gets a small scholarship from her work, so that needs to be accounted for. Sarah has incurred a HECS debt – how should that be managed?
What about the things that the family own or they owe money on? When we had a “reference dataset” it was a chore to think about what these might be, but once the dataset became a family, the job was much easier –both Sarah and Dave need a car – let’s let one of them have a loan, and the other one fully owned: there’s the house that they own, but a mortgage as well. Do they have shares or are they investing in Crypto? Is Sarah financing her studies using a loan? There is a credit card, which is always well used, and of course bank accounts.
What do we do with Dave and his plumbing business? Homefinance is not designed to deal with business accounts – it has no provision for calculating BAS or other taxes, or for debtors or creditors, or for invoices or for paying wages. So that is done by Dave’s accountant and book-keeper – all we see of Dave’s business is the drawings from the family trust.
Dave and his business is a good example of another advantage of using a reference family rather than an abstract data set. The family sets limits – when the plumbing business was developed, there were conversations about the role of Homefinance in the business. With an abstract dataset, adding on yet another function to Homefinance was logical, and could have led us down a very involved and unnecessary rabbit hole (or is it a python hole). When it became the Harris family, it was easy to quickly say this was a boundary Homefinance did not need to cross, because we were talking about how a family might organise their private finances. Homefinance needs to stick to its simple aim of assisting users to organise their domestic finances and make information available that would be useful to their professional financial advisors. Business finance needs a different set of tools, which are better kept in more business focussed software.
The Harris family began as a way of testing HomeFinance. Somewhere along the way, they became part of designing it. And, as we’re discovering, they’ve developed a rather inconvenient habit of leading us into places we hadn’t expected to go.
We’ll explore some of these places as we journey through the land of HomeFinance
