Do you have a project that is chronically late? Have endemic crunch times? On larger projects, it's common to find waterfall team dealing with creeping tech debt and a demoralised teams.
If you're looking for a way to bring your product back on track; review the below - focus on the eight great ideas and avoid ever using the four horrible ones.
This is paraphrased from: http://jockeholm.wordpress.com/2014/04/27/four-horrible-and-eight-good-ideas-in-a-project-crisis/
Four Horrible Ideas
1. Add more people ("resources")
From: http://en.wikipedia.org/wiki/Combinatorial_explosion#Communication
Brooks' Law indicates that adding more people to a late project makes the project later. On an early project, this may be considered, but take into account ramp up time to become productive and communication overheads.
Combinatorial explosion is the increase in communication lines as the number of nodes increase. Adding more people to a complex situation simply makes this phenomenon harder to manage. Implement regular communication times (such as standups) to ensure single-channel communications.
2. Delay the release
From: http://leanmanufacturingtools.org/732/muraunevennessproductionleveling/hockey-stick-graph/
Resorting to this obscures the real problem, which is one of prioritisation , inspection and adaptation to changing environments. Do not set up for failure by driving schedule as the only factor under control. The success of the product could be ensured via other mechanisms:
- consider releasing more frequently. Frequent releases lower the cost of not releasing; whilst providing overall agility - being able to respond to customer feedback is far more important than 'getting it right the first time'
- Consider prioritisation by value. Making your release plan more lean via product planning, will deliver more for effort, optimising that which is delivered.
- Fixed release dates have a way of creating the 'hockey stick'; a far more inefficient use of team effort. Instead if delivery dates are fixed, allow scope to flex in response.
3. Crunch mode
Avoiding the social and human cost of this is a social debt that employers and managers are obligated to mitigate on behalf of their staff. Crunch mode does not lead to more productive teams. Avoid this at all costs - use any of the great ideas below.
4. Lowering quality standards
Consider the cost to your tech-debt. It's hard for an organisation unfamiliar with the concept (and thus not making it visible) to come to grips with this building up. All too often, tech debt is added to over time under poor management, leading to ever slower projects, leading again to more tech debt being created as quality is continually sacrificed.
Eight Good ideas
1. Sit together
Co-location creates opportunities for high-bandwidth communications to occur between team members reducing combinatorial explosions. Add to the richness of body language, tone in voice and pitch, as well as ad-hoc use of tools such as standing in front of a whiteboard to illustrate - leverage the communication modes that our species evolved both physically and socially with.
2. Borrow specialists
In situations where a new technology is being used or major technical impediments exist, having an expert on devops, infrastructure, new technology or automation testing can make all the difference in productivity.
3. Move to weekly cycles
A shorter cycle leads to more frequent inspection of the work product and the manner that the work is done. It also tends to create more flow in the creation and delivery of work. Watch out for impediments against completing the cycle in a week, such as sluggish product leadership or trailing product validation (e.g. manual test phases).
4. Active product leadership
The most important idea here. The pillars of agile require "visibility, inspection and adaptation" on both teams/people and work products. What often drives chronically under-delivering projects is not the productivity of the development team - but the capability of designers and product leaders to make visible the status of the work product, inspect and adapt to it. Is the product as lean as it can be? Is the critical path to the product value (the minimum viable product) being created first? If not, then see the below step - use story mapping.
Koch Curve
5. Basic solutions first
The third most important, after story mapping. Don't over-design the solution - that represents a very expensive, wasteful and thoughtless way of developing products. Create large distinctions in your product - MVP being the most prominent grouping of value - and deliver that and only that first.
6. Use story mapping
The second most important - find out the stories - articulated as value, not as screens or other non-customer facing components - that comprise the minimum viable product. Save the articulation of lesser value for later.
7. Limit work in progress (WIP)
Do less at the same time to do more overall. Helps individuals and teams reduce waste by overburden; and create flow and productivity in context of the work they need to do.
8. Deliver frequently
Assuming that active product ownership exists, delivering to the customer frequently creates a cycle of product review with real customers that will encourage the product to evolve to identify the highest value to be iteratively and incrementally delivered.
