Three short questions and answers to finish off my series of left over questions about #NoProjects, #NoEstimates and the Continuous model.
Q4: How do we prioritize and organize requests on a product that are from opposite business owners? – for example legal (who wants to reduce the risk and annoy more customers) and sales (who want to increase the features and simplify life) can be arbitrated in a backlog?
You can think of this as “which is worth more apples or milk?” It is difficult to compare two things which are actually different. Yes they are both work requests – or fruit – and each can make a case but at the end of the day you can’t make everything number 1 priority.
In real life we solve this problem with money.
Walk into your local supermarket. Apples, oranges and milk are both price in the same currency, sterling for me, Francs for the person who asked this question, maybe Euro’s or Dollars for you. So if we can assign value points to each request we are half way to solving the problem.
Now sales will argue that without their request there is no real money so whatever they ask for is worth more. And legal will argue that nobody wants to go to jail so their request must be worth more. You can set your analyst to work to calculate a value but a) this will take time and b) even when they have an answer people will dispute it.
Therefore, I would estimate a value – planning poker style. With an estimates value there is no pretence of “right” or “correct”. Each party gives a position and a discussion follows. With luck the different sides converge, if they don’t then I average. Once all requests are valued you have a first cut at prioritisation.
Q5: How to evaluate the number of people you need to maintain software?
I don’t. This is a strategic decision.
Sure someone somewhere needs to decide how much capacity – often expressed as people – will be allocated to a particular activity but rather than base this on need I see this as another priority decision. If a piece of software is important to an organization then it deserves more maintenance, and if it is not important it deserves less.
You could look at the size of the backlog, or the rate of new requests and contrast this with the rate at which work gets done. This would allow you to come up with an estimate of how many people are needed to support a product. But where is the consideration of value?
Instead you say something like: “This product is a key part of our business but the days of big changes are gone. Therefore one person will be assigned to look after the software.”
If in three months more people in the business are demanding more changes to the software and you can see opportunities to extract more value – however you define value – then that decision might be revised. Maybe a second person is assigned.
Or maybe you decide that maintaining this product isn’t delivering more value so why bother? Reduce work to only that needed to keep it going.
Q6: How do you evaluate the fact that your application becomes twice as fast (or slower) when you add a new feature in a short period of time?
Answering this question requires that the team has a clearly defined idea of what value is. Does the organization value execution speed? Does the organization value up-time? Does the organization value capacity?
Hopefully some of this will have come out of the value estimation exercise in Q4, if not the analysis is just going to take a bit longer. The thing to remember is: what does the change do for the business/customers/clients? Being faster is no use in itself, but doing X faster can be valuable.
The real problem here is time. Some changes lead to improvements which can be instantly measured. But there are plenty of changes where the improvements take time to show benefit. Here you might need to rely on qualitative feedback in the short run (“Sam says it is easier to use because it is faster”). Still I would keep trying to evaluate what happens and see if you can make some quantitive assessment later.
Notice that Q4 and Q6 are closely related. If you have a clear understanding of why you are doing something (Q4) then it becomes easier to tell if you have delivered the expected value (Q6). And in trying to understand what value you have delivered then you refine your thinking about the value you might deliver with future work.
Another feedback cycle.
These questions concludes the series of question carried over from the #NoEstimates/#NoProjects workshop in Zurich – see also How should we organize our teams? – Dealing with unplanned but urgent work – How do we organise with a parallel team? – if you would like me to answer your question in this blog then please just e-mail me.