I’m just about to start a new project, finally. It’s
starting a good three months after it was supposed to due to client
procrastination but with the same goals and the same end date which will make
for a challenging delivery. You’ve heard this before, right? Throw in some off
shore testing resources and the possibility that some team members may have to
spend up to seven days in swine flu quarantine at home during the project and
the potential for failure is high. With this in mind we decided that an online,
collaborative agile management tool would be essential.
Turning to the internet, we quickly discovered that there was
no definitive choice of agile process management product to meet our needs. They all seem to be overly complicated,
trying to do everything and be ultimately flexible or too simplistic, unable to
do enough. Nothing seemed to toe the middle ground. Once we’d established that
the solution wasn’t going to be simple or obvious, we set about being a little bit
scientific about our search. We thought about the process of developing in
agile, mapped the key interaction points in the process and turned these into judgement
criteria. These were:
- Backlog – Can you create
and manage a product backlog of stories simply and effectively, prioritising
and ordering them?
- Estimating – Can you estimate
the stories on your backlog, simply, in bulk after the backlog is created?
- Stories – How expansive
are the stories and can you nest them within epics or turn them into epics
at a later date?
- Tasks – Can you create nested
tasks under a story to represent the actual work that needs to be done,
for more accurate tracking?
- Testing – Does the tool
allow for special testing tasks and testing oriented views of progress?
- Teams – Can you create
teams of people and manage their capacity over time?
- Planning – How helpful is
the tool in planning sprints and matching selected stories to sprint
capacities. For a bonus, are these capacities dictated by the assigned
team capacity?
- Progress – How easy is it
to update progress on tasks and stories and does it track only how long is
left on a task or how long was spent as well?
- Board – Does the tool
provide a story board or a task board (or both) and how interactive is it?
Can you drag and drop to update tasks, can you take ownership of a task
directly from the board?
- Burndown – How clear is
the burndown and what other techniques of tracking progress exist?
- Usability – Generally, how
easy to use is the tool?
These criteria should describe the things that most agile
teams would need in a product but in order to ensure we obtained the best for
us, we then weighted them as follows:
- Backlog – 10 (This is
essential)
- Estimating – 6 (Need it
but, we can cope with a little pain up front)
- Stories – 10 (Another
essential)
- Tasks – 4 (We can manage
this with more granular stories if we have to)
- Testing – 2 (Not saying it’s
not important but, differentiating testing tasks from other tasks isn’t.
We’re happy to manage this with standard tasks or even stories)
- Teams – 2 (Not vital, we
have a short project and a small team)
- Planning – 8 (Not
essential but very useful and speeds up sprint planning)
- Progress – 10 (The crux of
everything really. If this isn’t right, nothing upstream or downstream is
useful)
- Board – 10 (This was the
need that started the search and this visualisation remains an essential
for us)
- Burndown – 6 (The defacto
view of project progress perhaps but not all that vital on a project as
small as ours)
- Usability – 7 (Don’t want
it too clunky but, we’re all intelligent guys and we’ll figure it out)
The final task was to build a shortlist of candidates. Inevitably,
we had to include a cost factor. Our project is not huge and, we can’t say with
any certainty that whichever tool we choose will be used beyond its end. So, we
worked out comparative costs for 10 users over 6 months and excluded the most
expensive. A couple of others were excluded as they only worked on MS SQL which
didn’t work for us in a Linux and Oracle environment. The final list is below,
and I’ve also listed those that didn’t quite make it.
The following were looked into but excluded due to cost:
The following were excluded due to technology constraints:
Now all that was left was the scoring. We downloaded or
signed up for the free trials of each product and set about inputting our data
and running a mock project for a few days. This gave us a reasonably real-world
view of the products’ performance. Fed the scoring into a spreadsheet and came
up with some results. As this explanation has become rather long I will post
the results and a short review of the products in the next blog.