Planning Software Activities – A tools perspective

4 minute read
What is DevOps?

What is DevOps?

DevOps is the continuous delivery of prioritised value to an organisation. For more click here.

Background

Setting and meeting expectations matters … a lot! It’s how we build trust e.g. doing the things we said we would. Businesses will always need to prioritise feature delivery and assign dates to their delivery. Be it competitive or straight forward revenue pressures very few businesses can survive without publishing a roadmap of some sort. Atlassian themselves uses their own product so called Advanced Roadmaps (AR) and this is how they present their view to their customers.

AR has its genesis In 2014 when Atlassian launched Jira Portfolio but more recently as they migrated core products to the cloud they relabelled the product Advanced Roadmaps and it now ships with the Jira Software Premium edition.

Atlassian also acquired (2020) and launched Jira Align as an Enterprise scale Agile planning solution that is aimed at very large organisations with thousands of engineers and a C suite that wants dashboards rolling from Jira Software issues into strategic objectives. This link on Atlassian’s site compares and contrasts Advanced Roadmaps and Jira Align.

In this blog we look at Advanced Roadmaps to help organisations make informed decisions about their tooling approach for medium (months) and long term planning (up to 2 years) and to help determine if the Standard or Premium edition of Jira Software is worth the extra spend.

(There are other reasons why organisations will purchase Jira Software Premium over Jira Software Standard but Advanced Roadmaps is one of the major functional differences.)

What are the use cases for Advanced Roadmaps (AR)?

Apart from setting and meeting expectations, trade offs are the other significant driver. Requirements change as businesses execute a strategy and re-prioritisation of feature backlogs demands that you model different scenarios with stakeholders thereby informing future decisions.

Two key Advanced Roadmaps concepts

1. At the heart of this product is the concept that there is at least one constraint any planner will have to contend with. Reality is that you can only change any 2 of the parameters in the Iron Triangle to solve the constraint that is the 3rd. If you are time constrained to a release date you can either reduce scope or add resources. Or if you have no time limit the scope will be bounded by the available resources. If you don’t have any of these constraints you may not need AR and can I work for you? 🙂

2. The other major point to remember is that Jira Software Premium ships with a Sandbox. This means that you can ingest your prod Jira Software issues and play with that data but you only write back to your production Jira Software when you are committed to a new plan. Clearly the Sandbox also has broader usage as a test bed but as you will see below scenario modelling is best done in a Sandbox.

As compared with Basic Roadmaps which ships with the Standard edition of Jira Software, Advanced Roadmaps gives several additional layers in the hierarchy and its useful to note you can label them whatever you like. AR also provides dependency management between projects, scenario planning and auto scheduling.

Auto Scheduling

This is a useful feature of the tool. The inputs are;

Sprint Assignments – where issues have been explicitly assigned
Estimation method – story points or hours
Start and end dates for releases (releases and versions are synonymous)
Ranking of issues – position in backlog
Team members – offers the possibility to add TBH or unassigned engineers
Team’s work e.g. sprint duration or Kanban
Dependencies (see below)

A key point to note is that the Auto Schedule output can be viewed as a start point for manual overwrites. It uses AI to populate a plan and indicates if all of the scope can be delivered (Green) or if more resources/time is required (Red). Think of this as a quick sanity check for your as-is position.

Dependendencies

It’s rare that a dependency is singular. Invariably dependencies in software development are not just code dependent but also dependent on for example the availability of a specific engineer/team who understands a specific piece of code. Advanced Roadmaps accounts for these inter-linked issues as an input to Auto Schedule.

Scenario Planning.

There are two options for any scenario based on the Iron Triangle

Release date is fixed
Release date is dynamic

This is a ‘what if’ tool as you take an auto schedule output and for example remove a feature or add more scope or add more resources. This is ideal for worst and best case planning as you engage stakeholders prior to ‘committing’ back the new schedule to your prod Jira Software instance. Another useful aspect of this tool is the ability to look at cross project releases where different teams will be releasing together albeit there are no dependencies.

So much for the theory – how does it work in practice?

An end users story after 6 months of usage

This global client has circa 1000 engineers in 40 teams and the company has grown through acquisition in the last 3 years and now has a total of 10 products grouped into 3 three strategies. (see diagram above)

The problems facing this business were;

1.Map product interdependencies … but not necessarily at the Jira ticket level
2.Ensure engineers are connected to a higher purpose.
3.Get visibility of all work
4.Identify duplications or work which cannot be connected to value

Engineering teams are in the process of connecting their Jira issues to the Product Roadmap initiative. This is something they want to do so that item 2 is more comprehensively addressed.

Auto-schedule is not a feature this client has plans to implement. They do use story points but do not believe their velocity is predictable or that their estimation is reliable when undertaken manually. Therefore they have concluded as they do not fully understand the science that underpins auto schedule and until the teams are more completely connected to AR auto-schedule can wait.

The client did hit an upper limit of 5k for ingestion of issues.

They evaluated Aha, Dragonboat, Product Plan and Product Board – and there was no price difference between. They had a small number of Aha licences but found the learning curve was quite steep as compared to AR. The client then realised that for other reasons they had licensed Jira Software Premium and as cost became a consideration they felt the implementation was much more straightforward and it was effectively ‘free’.

The client believes they are probably only using 30% of AR’s capability but because they have a much deeper hierarchy they can finally get their arms around their product strategy and scope of all the work. This has been of huge value to the business


POSTED IN: Agile, DevOps process, DevOps tools