[Tl;Dr: The essence of software and product development is about continuously exploring and meeting folks’ needs. How about as in industry we stop shilly-shallying and place this reality at the core of our approach to development?]
The Essence Of Software And Product Development
Essentially, every approach to software and product development is about a flow of things (including services) whereby a development team, group or organisation attempts to satisfy some selection of folks’ unmet needs.
In batch-and-queue (e.g. Waterfall, etc.) approaches, needs are batched-up into large batches, and flow (not very effectively) through a sequence of queues. Each batch (in the pathological case, only one) eventually gets dropped onto those assumed to have said needs, and that’s about it.
In iterative (e.g. Scrum, Kanban, etc.) approaches, needs are batched up into many, relatively smaller, batches, and flow (rather more effectively) through a sequence of queues (at its simplest, e.g. Backlog, Doing, Done). Each batch gets dropped onto the folks assumed to have said needs, and these folks get the chance to say if their needs have been met well enough, or not. When budgets and timescales allow, a later iteration can then have another go at better meeting those needs deemed to have been poorly-met, along with a new attempt to address any other priority needs that might have emerged along the way.
Different approaches differentiate themselves on e.g. batch sizes, selecting of the kinds of folks whose needs will be attended to, cadence and strategies.
Needs Trump Value
Why Needs flow? Why not just stick with the more common “value” flow? Well, for me there’s a whole bunch of issues around the notion of value. Including:
- Value for whom?
- Who gets to specify how any particular “value” is measured or quantified (most often folks ill-equippped to do so).
- The whole notion of “value” (often overlooked, often misunderstood).
- Value is too often though of in simple economic or financial terms. Human beings perceive value differently, and much more richly (e.g. emotionally).
- How can we tell when we’ve delivered “value”?
So, I propose that needs trump value. And therefore, that when focussing on flow, it’s more useful – and effective – to focus on the flow of needs.
The above diagram shows a system boundary – this could be the whole organisation, the development organisation, or just one development team. Bundles of external needs flow into the system, and merge with other bundles of needs coming from inside the system. These bundles pass though various stages (and queues) where “development” strategies are uses in moving the bundles from stage to stage, and from queue to queue. Ultimately, candidates (things we hope will meet the needs that entered the pipeline) flow back to external parties, and also to internal stakeholders, such as the developers themselves.
This perspective allows for the simple integration of dealing with impediments. Things that are impeding e.g. flow or the outflow of effective candidates may themselves be needs. And may enter the pipeline much as any other needs. Of course, there may be some( or many) such impediments than no one has any need of dealing with. These may be filed or ignored.
Whilst FlowChain was conceived to be awesomely effective at handling bundles of e.g. User Stories, Use Cases, Improvement Stories, etc., it can very easily embrace the unit of flow being bundles of needs, or – in the single-piece-continuous-flow scenario – individual needs.
Limiting WIP, Making Things Visible
From the diagram, you may begin to see how we might simply make the flow of needs visible, should we so choose. And personally, I’d like to make visible the ultimate fate of the candidates too, with a little expansion of the scope of the diagram.
And the queues (particularly the first) can serve as a convenient means to limit Work In Progress (a.k.a. Needs in Progress).
It may not be immediately obvious, but the NeedsFlow perspective implicitly removes the need for projects. NeedsFlow decouples development (flow) from products and product lines (each of which meet a more or less related collection of needs).
I disagree with your assertion that “The essence of software and product development is about continuously exploring and meeting folks’ needs.” The essence of software and product development is, well, software and product development. If all folks’ needs can be attended to 24 X 7 while their doing what they’re paid well to do, then that’s terrific; icing on the cake. I’m a knowledge worker and I don’t require my needs to be attended to constantly. I think it would be selfish and impractical of me to do so. But that’s simply my opinion.
I guess you find the idea of indulging developers (a.k.a. knowledge workers) somewhat… disconcerting?
No, I don’t find it disconcerting. I enjoy indulging fellow developers very much. I do it daily with my colleagues and my bosses. But it’s not all peaches and cream, all the time. Passions flare, and tensions ebb/flow in the workplace; just like outside of the workplace – real life.
I am at a loss, then, to understand your original comment.
As usual, we have a huge communication gap.
Pingback: Five Blogs – 1 April 2015 | 5blogs