# Sunday, 13 December 2015

A living backlog needs a strong internal structure (skeleton) in order to effectively guide a project and maximise the value delivered to users.

That structure enables the backlog’s primary purpose of alignment to the business with a clear sense of priority and delivery sequence.

  • Time and priority to know when value is likely to be realised and the relative priority of work. Using Iteration Paths
  • Business-centric to enable stakeholders to guide development and understand the evolving scope. Using Area Paths

A ‘living backlog’ is one that:

  • Guides decision making, identifies risks and possible opportunities
  • Is comprehensible to all stakeholders and is ‘prioritise-able’
  • Provides a solid basis for metrics to answer the question ‘are we on track?’
  • Is up-to-date and can incorporate emergent changes in scope
  • Reflects the different streams of work including: planned (project), BAU, Support and non-functional work

As opposed to many of the backlogs I see, which feel more like a necessary but unloved artefact to be dragged along by the team.

The four dimensions to structure and slice a backlog in TFS are:

  • Iteration paths to provide a time and priority-based view
  • Area paths to provide the business with a recognisable perspective that will inform decisions
  • The work-item hierarchy (e.g. Goals-Features-PBIs) to align to the business case
  • Tags to flexibly ‘cross-cut’ the backlog thematically

 

Iteration paths

The Iteration path enables the creation of burn-up chart for the next release / deployment and therefore key to be able to answer the key question:  Are we on track?

Well planned and maintained Iteration paths enable a simple burn-up chart like this.

burnup

This chart is from our www.senseadapt.com collection of simple visualisations for TFS. For more see my previous blog on the 'Landing Zone' chart

Deployments to live might be a regular cadence of monthly /quarterly drops, with a preference for more frequent, smaller batches.

Three ways to use Iteration Paths

Kanban

In a Kanban-like approach each Iteration represents a ‘bucket’ of functionality to deployed.

  1. Release 15.8
    1. PBI_1 to PBI_n
  2. Release 15.9
    1. PBI_1 to PBI_n

Scrum

Teams using Scrum will include the sprint iterations within the iteration for the release/deployment

  1. Release 15.8
    1. Sprint 1
      1. PBI_1 to PBI_n
    2. Sprint 2
      1. PBI_1 to PBI_n
    3. Sprint n
  2. Release 15.9
    1. Sprint 1
    2. Sprint 2

Mature Agile

Teams who deploy frequently will benefit by using the Iteration path to manage their different release horizons, for example:

  1. In the next month
    1. PBI_1 to PBI_n
  2. Within 3 months
    1. PBI_1 to PBI_n
  3. 3-9 months
    1. PBI_1 to PBI_n
  4. Probably never
    1. All the rest of the PBIs

The ‘probably never’ iteration is a useful parking spot for PBIs that are unlikely to see the light of day – but which stakeholders get twitchy about deleting. This keeps them out of the metrics and maintain focus on the more valuable PBIs

Work-item Hierarchy: Goals – Epics – PBIs – Tasks

These can be used to align the backlog to the original business case

  • Goal: the business or technical goal that we are trying to achieve
  • Feature: large stories or logical collections of PBIs.
  • PBI: (Product Backlog Item) – the value-carrying piece of work that will be tracked through to live.
  • Bugs: can either be linked to PBI or Epic
  • Tasks: The actual work that needs to be done to deliver a PBI (mainly used in Scrum)
  • Tests: linked to a PBI or Bug

workitem

Note: Your work-items may have different names from those above.

Area Paths

A living backlog must actively engage business stakeholders and the Area Path is the best way to do this.

Area Paths provide a recognisable, business-centric structure for the backlog. The structure will vary by type of project, see the three examples below.

Talk through the best structure with business stakeholders, as this is their window into the project.

The Area Path branches enable rolling up of PBI data. A simple way to show the relative effort and progress for a particular aspect of the project.

Area path by stakeholder

In the example below the money to fund the teams comes from the business divisions. The HR stakeholder naturally wants to quickly see a summary, including progress, of all the work that she is funding.

There will also be non-functional work e.g. automation of testing and deployment, which should have its own Area Path, if it does not relate to a specific top-level stakeholder.

  1. Marketing
    1. CRM
      1. Campaign content
      2. Campaigns
    2. Planning
    3. Tracking
    4. Advertising resources
  2. HR
    1. Recruitment
    2. Reviews
      1. Permanent
      2. Contractor
    3. BAU
  3. Sales
    1. BAU
    2. CRM
      1. Prospects
      2. Clients
      3. Brokers
  4. Non functional
    1. Automation
      1. Testing
      2. Build & deployment
    2. Refactoring
      1. Decommissioning system x
    3. New API for brokers

Area path by user / role

In this example from an online retail application, the Area Path is used to slice the backlog by role.

  1. Marketing Manager
    1. Planning and reporting
    2. Managing range
    3. Managing Product
      1. Import new products
      2. Archive older products
    4. Promoting
      1. Promote a range
      2. Couponing set-up
    5. Managing content
  2. Customer delivery assistant
    1. Preparing for delivery
    2. Delivering customer orders
      1. Taking payment at door
      2. Getting customer feedback
    3. Returning to depot

Area path by functionality

Here the Area path is used to slice up a product like Outlook, by functionality

  1. Organise Email
    1. Search Email
    2. File Emails
  2. Manage Email
    1. Compose email
    2. Read email
    3. Delete email
  3. Manage Calendar
    1. View calendar
    2. Create appointment
  4. Manage Contacts

Tags

Tags are great at thematically ‘cross-cutting’ the backlog.

For example:  all PBIs that relate to a compliance strand of work; PBIs that use a specific technology; PBIs that have dependency on a third party service

A word of caution:

  • Tags are not easily delete-able. You need to first remove them from every item, then wait a few days for them to be cleaned out of the database. See this description of tag-cleansing process 
  • it’s easy to duplicate or slightly misspell Tags, leading to consistency problems

 

Use a ‘Two Dimensional’ backlog

A ONE dimensional backlog is a simple, powerful way to prioritise work however they are tedious to review and difficult to comprehend.

A poorly understood backlog makes it difficult to see opportunities and deliver more value from the project.

A TWO dimensional backlog is simple way to show both the time aspect of the project and the functional / business perspective.

  • Time (iteration paths) are shown down the side
  • Functional / business (area paths) are shown along the top

2D backlog

Add some colour and shapes and it becomes multi-dimensional. The chart above uses colour to show the state of the PBI.

Below the size of the pie chart shows how much work is in each area whilst the colours show the state.

Backlog by scope

The two dimensional backlog is analogous to ‘Story Mapping’, a popular technique for gathering requirements for Agile projects. See http://jpattonassociates.com/user-story-mapping/

I urge to read Jeff’s book, it’s a great summary of better ways to gather requirements.

Later blogs will look at monitoring the health of the backlog, what are the vital signs that you should be aware of.

.

Tags:

Sunday, 13 December 2015 12:07:00 (GMT Standard Time, UTC+00:00)  #    Comments [0]


# Wednesday, 30 September 2015

Are we really 15 years into Agile and still failing to master one of the foundational principles? That is to use data to provide frequent, actionable feedback loops. Most teams still rely on the most rudimentary of metrics… velocity and burndowns.

We know we can only deal with the complexity of software development by using, ‘inspect and adapt’ cycles. We have collected lakes of data in every TFS, Jira and Rally instance and yet most of it remains as clear and useful as a stagnant pond.

Metrics are tainted by years of association with ‘governance’ and of being used to feed the ‘process Gods’. The idea that the individuals within teams might actually benefit from this insight seems to have been overlooked.

SenseAdapt is a collection of simple visualisations that enable teams and stakeholders to sense what is going on in their system and adapt it, in order to deliver more value.

Our aim with SenseAdapt is to bring objectivity and help create a shared, common understanding of what is going on. The conscious and sub-conscious motivations of team members, project managers and stakeholders too often clouds the picture.

The visuals are being used by three of our largest clients and continue to evolve rapidly with their feedback.

Data-based, Actionable Insight

  • Data-based
    • We utilise the data that teams are already collecting. Although, now that that data benefits them, there is more reason for them to care about its quality
  • Actionable
    • Just one click away, or on a dashboard, and displaying the real-time data
  • Insight
    • Simple, comprehensible visual representations
    • Relevant as it shows how the system is behaving right now, not just retrospectively

Note: SenseAdapt continues to evolve. Please get in touch if you would like a demo or join our other clients in trying it on site. For the moment it is free – in exchange for your feedback.

 

Team-Facing Visuals

Used through the day, possibly in a dashboard, and draw attention to the parts of the system that need action. For example; PBIs (Product Backlog Items / user stories) that have got stuck or forgotten, the levels of WIP per person, the number of unestimated or PBIs that are too big.

The visuals also provide useful input into the retrospectives.

WIP by person

Highlights who is overburdened in the team and therefore needs to complete a few items of work – rather than continue to start new items.. i.e. adopt the ‘Stop starting, start finishing’ mantra. Could be used at the daily stand-up

WIP

 

Cumulative Flow Diagram (CFD)

Shows the behaviour of the system over time, the bottlenecks, levels of WIP and lead times.

cfd

 

Requirements readiness

Draws attention to the PBIs that need to be refined, sliced smaller or estimated. The chart shows the numbers of PBIs of each size, and their state. It clearly shows the unestimated stories as well.

ReqReadiness

 

Sprint (Iteration) burndown

This is better than the commonly used sprint burndown, as it shows the rate that the team is actually completing stories – not just tasks.

This example shows a typical anti-pattern of last minute closing of stories in a sprint. The story burndown, in points, is shown on the right hand scale, with task burndown on the left.

The use of live data means that this chart can show an ‘intraday’ burndown

SprintBurnDown

 

Backlog health

Shows whether PBIs are in the correct Iteration Path or Area Path (when using TFS).

Most metrics are driven from the Iteration path, therefore PBIs in the wrong iteration can lead to a misleading picture and incorrect forecasts.

BacklogHealthIteration

This is toggle-able between Iteration paths (above) and area paths (below)

BacklogHealthArea

 

PBI Staleness

Teams quite often come to overlook PBIs, which may stall in a particular state or just be forgotten, this particularly the case if there is a lot of WIP (Work in Progress).

This chart draws their attention to PBIs that have not been amended for and need some attention.

PBI_Staleness

Business-Facing Visuals

These charts help stakeholders understand where the project is and when or what options there are for intervention.

Live project data – one click away. Each saved chart has its own URL that can be emailed or held in a project page or wiki.

They help to keep stakeholder expectations realistic and aligned. Significantly more informative and transparent that opaque RAG updates, which hide as much as they reveal and remain largely subjective.

Burn-up / Landing Zone

The burn-up clearly shows the range of likely delivery dates of a project. The example below is from a client with two teams totalling 20 people.

The project is likely (we deal in probabilities) to land in the shaded green ‘landing zone’. It’s clear, in this example, that one of the key risks to the delivery date is the rate of increase of the scope.

o clip_image008

This forecasting is based on standard deviations from the mean , which can be varied by the user, using more or less history in the calculation of the mean and choosing how many standard deviations for the forecast.

Most teams have settled for burndowns and velocities which offer far less insight and awareness of possibilities. See my other post for more on the Landing Zone

In the near future we will use Monte Carlo (see below) simulation to model and forecast both the scope and the throughput rates.

Feature Progress

Shows the progress of PBIs within a feature. Business stakeholders can clearly see the relative size of features, as well as the progress.

FeatureProgress

Release Treemap

Shows the status and relative size of all features. Clicking on any feature will show the same for its child PBIs. Features and PBIs can be opened directly in TFS.

TreemapPBI

 

Simulating the future

Thought leaders, like Troy Magennis, toil on the edge of our Agile consciousness, promoting ideas that we should all have embraced in a heartbeat – specifically that the data of how our systems have behaved to date is the best predictor of how our systems will behave in the future.

His website and blog Focused Objective makes a very strong case for this approach.

Essentially, if you want to predict when a project is likely to land, you need to:

  1. Create a mathematical model of how your system, whether a single team or a programme, has been behaving historically, based on the actual data
  2. Then take the remainder of your backlog and run it through that model thousands of times, giving you a frequency distribution of the possible dates it could finish.
  3. Use that range of dates to engage stakeholders and start discussions about what could be done to land the project successfully.

After 10,000 runs the output looks something like this.

image

We will write more extensively about the use of simulation within SenseAdapt in later blogs.

.

Tags:

Wednesday, 30 September 2015 16:15:29 (GMT Daylight Time, UTC+01:00)  #    Comments [0]


# Thursday, 14 May 2015

I’m embarrassed to admit that, as a long time Agile coach, I hadn’t appreciated the benefits of burn-up charts over burn-downs, merely seeing it as a mirror image - I was wrong.

However it has brought me to this metaphor of landing the project in a zone where two joysticks, held by different stakeholders, guide in the project (or MVP or ‘next release’). The PO (business) holds the scope ‘joystick’ whilst the team adjusts the ‘throughput rate’.

Being co-pilots in this joint endeavour helps foster collaboration between the two ‘sides’.

i find that the most insight-generating metaphors are simple and incorporate a familiar aspect of our physical world.

Make the most of the historical data we have

Our historical data of ‘completed work’ shows the rate of delivery (velocity). Which we can extrapolate into the future using pessimistic and optimistic rates.

A vertical line down from where these cross the scope line, gives us a date range that this project will finish. In this case I’ve assumed the scope doesn’t change after today.

landingzone

The angle of those pessimistic and optimistic throughput rates is largely determined by the team and overall system effectiveness.

Note: We currently use simple calculations to decide the angle, e.g. optimistic (median +25%) and pessimistic (median – 25%). Alternatively a moving weighted average works well too.

Not forgetting the ‘S curve’ that David Anderson writes about here

This shows there is often lowering of the throughput rate in the last 20% of a project – assuming that the definition of done was not quite perfect – which is common!

s curve

 

We are also developing projections based on ‘probabilistic forecasting’ and Monte Carlo simulations to provide more clarity on the likely outcomes. Troy Magennis has been leading this approach in the Agile world – this video is a great introduction

Step 2 – Scope evolves

We know the scope line will probably continue to drift upwards, in fact there are probably two future scope lines:

  1. The desired scope that was in the business plan and stakeholders thought was necessary when the project was first conceived.
  2. The minimal scope, or MVP, that would deliver value to users, generate learning and is a viable delivery.

landingzone1

We need to develop better ways to model this change in scope, however we’ve all seen this in projects, even if the root causes vary.

There are two possible date ranges within which this project will go live.

  • Minimal scope = mid-June to early-August
  • Desired scope = mid-July to mid-Sept

 

The Landing Zone

It’s easier to visualise and to say, with a fair degree of confidence, that the project is going to land somewhere in this green zone.

landingzone2

Representing the truth in this comprehensible, palatable way stimulates many more possible interventions, for the team and business, than a RAG status ever could!!

 

Options to hit an immovable date

To be confident that we can launch at that conference in mid-August, we can see what options we have to reduce scope to come within the pessimistic line.

It’s also clear that we will at least have the MVP by the conference and possibly a little more.

landingzone3



.

Thursday, 14 May 2015 15:40:00 (GMT Daylight Time, UTC+01:00)  #    Comments [0]


# Monday, 16 June 2014

A client recently told me of some members of a development team saying how they couldn’t see the value of adopting Agile. It’s not uncommon to hear similar sentiments in many teams.

This got me thinking about how we often externalise the indefinable in order to; understand it, sell it, master it and sometimes resist it.

The common coaching attempt to reposition Scrum/Kanban as a ‘framework’ rather than a ‘process’ doesn’t come close to the visceral resistance (fear) we feel to something ‘other’, that is outside of us.

The real challenge is not one of semantic classification, but of human self-identification and engagement.

In essence this is about our ‘way-of-working’, something we are fully immersed in. Like being in water, we can’t push against it. This is what we do, hour-to-hour – it’s not external to us… it is us. And that’s the quasi-Zen bit**.

That Kanban board over there is merely a reflection of the way we currently work – it’s not ‘a process’. If it’s not an accurate representation – let’s improve it.

If it turns out that my way of working is less than perfect, and I find myself wasting time then what could I do to improve the situation. [Hopefully we move to replacing ‘I’ with a team ‘we’… but ‘I’ is where it begins.]

When it comes to Agile principles there is no external 'process' unless you choose to define one. Scrum, Kanban or XP are merely instantiations of those principles – they bring those principles to life.

We need to find a way for people to see beyond their own work and to the results of their collective efforts with others. After all, the ultimate purpose in our work is in delivering value to users – and that usually requires other people.

I’m keen on ‘optimising the whole’ and ‘Systems Thinking’, however that too is a form of externalisation and should be introduced carefully.

My current (June 2014) , evolving recipe for Agile adoption is:

  • Blend Agile principles into your existing way-of-working, at a rate they deliver measurable improvements and are ‘digestible’ by the individual and organisation
  • Catalyse the evolutionary process with a sense of urgency and in a clear direction – the ‘True North’ vision
  • Whilst keeping it safe to learn and recognising those that share their learnings.

And what I would say to a team is:

  • ‘Let us try to improve the way-we- work in order to reduce waste and frustration. ‘

This probably comes across as a negative perspective – and yet, I think it evokes a particularly powerful emotion in us – because it’s based on actual feelings.

A more ‘positive’ perspective like, “deliver more value” arises from abstraction that doesn’t resonate as deeply or evoke emotion in me. As an individual I can’t influence (on my own) the behaviour of that system – and so, whilst I might recognise it as a laudable goal, it doesn’t impel me to action – at least in the early stages.

Whilst there is positive emotion associated with recognition of a job well done or in the sense of completion when one puts something in the ‘Done’ column, I think the ‘negative’ is more likely to stop us tolerating the status quo.

 

** I am in no ways a Zen practitioner, however I did find these ten Zen principles have some resonance with Agile principles: Awareness, Present Moment Focus, Non-judgment, Acceptance, Validation, Tolerance, Compassion, Invitation, Patience and Practice http://www.drnataliemasson.com/TenPrinciples.html

Apologies if by the end of this you saw it as a thinly veiled attempt to blog a catchy title (See Zen And the Art of Motorcycle Maintenance http://en.wikipedia.org/wiki/Zen_and_the_Art_of_Motorcycle_Maintenance )  However, there is something illusive about this topic which requires a certain contemplative mind-set, which is another aspect to that I was trying to convey.

.

Tags:

Monday, 16 June 2014 15:32:37 (GMT Daylight Time, UTC+01:00)  #    Comments [0]


# Friday, 08 February 2013

Portfolio-level prioritising is fraught with problems… ambitious senior stakeholders, promoting their agenda, rarely meeting together to see how their projects compare to others resulting in a portfolio in continual flux – and not a little distrust.

Such vague, contradictory, fluid priorities at the top trickle down through the whole organisation… generating waste as they go….

Getting a stable, visible and well prioritised portfolio backlog is a key step towards successful Agile adoption.

Prioritising by ROI or a strict financial measure does not work – even though we wish it did and keep trying to make it so – an army of commercial accountants couldn’t set enough rules and then get everyone to play by them.

The underlying challenge

The real challenge is not whether to do a project – but which one to do next.

i.e. this is about prioritising, not about getting a business case signed off.

Every client I’ve worked with in the last seven years has had more than two years of projects with over 100% ROI in their backlog i.e. they would all pass the business case test.

The other challenge is to keep the business focused on the most valuable projects – and minimising Work In Progress (WIP).  The easiest way to keep insistent stakeholders at bay is to at least START their project… spinning up more and more WIP as we do so.   

Value is not enough

The value of an Epic (slice of project) or MMF (Minimal Marketable Feature) itself is not enough. Some valuable work is trivial to deliver, whilst another is costly – so our approach needs to quickly identify those with a high value:cost ratio – and get it moving through the system..

The value / cost ratio is the key– and we decide what constitutes value and cost – which means we can include some less tangible elements, like risk. We can also find a way to include ‘gut instinct’, which factors in our experience.

The ranking needs to be simple, quick, objective (as possible) and transparent.

The process itself adds value

The process encourages conversation and adds insight and understanding between stakeholders – this is not the wasteful process that some have suggested.

This approach helps to ensure the conversation at least starts from the same place – i.e. with a common understanding of what we are trying to prioritise.

No business should make a decision purely on such numbers – however, this weighting process at least allows for comparisons to be made.

How to get to a usable value / cost ratio

Both value and cost are made up of several elements, customised to the specifics of a company or programme.

For example, one of clients use the following:

image

These are then weighted by the stakeholders, depending on importance , with weightings adding up to 100.

Each element is given value of; 0,1,3,5 or 7, writing the text in the boxes is another valuable process for stakeholders to discuss and agree on.

This element score is then multiplied by the weighting to give an overall value score for that Epic or MMF.

image

For example:

Epic 1 ‘Remove step in the basket checkout journey’ scores;

  • 3 for ‘strategic alignment’,
  • 7 for ‘customer experience’
  • 5 for ‘cost of delay’

giving an overall value score of 3x65 + 7x 35 + 5x10= 440

Whilst the cost is:

image

Epic 1 scores:

  • 5 for effort to develop
  • 1 for technical risk
  • And 0 for architectural

Therefore Cost = 5x50+1x30+0x20 = 280

Giving an overall value / cost ratio of 1.57

All the epics can be summarised in a table like this..

image

At a glance we can see that Epic 3 gives us a greater return than Epic 1 – even though 1 has an overall higher value. This data is easily presented and understood by senior stakeholders.

It’s easy for one stakeholder to see what rankings have been given for an epic and to drill into the assumptions behind it and so we get discussion, trust building and a better decision.

The benefits are clear:
  • The assumptions being made about value and cost are made explicit
    • Asking why someone gave a 5 rather than a 3 quickly surfaces the underlying assumptions – and starts a valuable conversation.
  • More visibility, and multiple perspectives means better decisions and stronger buy-in to those decisions – and a more stable backlog
  • It’s quick – and there’s room for valuable instinctive ‘gut feel’ judgements
    • Rather than in overly (misleadingly) precise financial models
  • It results in a single number that can be used for an initial prioritising of the backlog
    • Stakeholders can then adjust the order for dependency, or other business reasons that may not be in the model.
  • Coming up with model is itself an useful step for senior stakeholders –
    • What are really the key components of value and their relative importance
    • Likewise for cost, there is a now a way to factor in risk and architectural impact and dependency and not just effort.

A way to calculate value delivered – another illusive metric

Having derived a value of an Epic it is now possible to begin tackle one of the most difficult questions facing an organisation or programme:

How much value have you delivered in the last quarter?

It’s now possible to total up all the epics delivered and their accumulated value and follow the trends… It’s true we don’t have an absolute value – but the trend does tell us something useful..

Note : Typically we have used to prioritise the MMFs and ‘Epics’ (slices of project) whilst in principle it could be used finer-grained user stories, the process would probably need adapting.

.

Tags:

Friday, 08 February 2013 13:42:07 (GMT Standard Time, UTC+00:00)  #    Comments [0]


# Friday, 23 November 2012

The simplest illustration of the benefits of focus must be this one – clearly illustrating that if we focus on _and finish_ one feature (or project) at a time then all 3 features are delivered quicker.

Agile emphasises delivering value sooner – in order to generate revenue, create lessons and validate assumptions.

clip_image002

  1. In the ‘current situation’ all three features are started as soon as they are requested – but none are finished until near the end.
  2. Focusing on one at a time means that A and B are delivered value sooner in the second example
  3. However, many studies (e.g. in Tom DeMarco and Tim Lister’s book ‘PeopleWare’) show that there is a real cost in switching between tasks.
    Therefore, assuming a saving of only 10% – we see that all three features are delivered significantly sooner. This is not just a marginal gain, it could translate into weeks of additional sales.

As well as the benefit of earlier value and lessons – don’t underestimate the energising effect on a team that delivering A, and knowing it’s in the hands of users, can bring. Focus increases the ‘realness’ of the work and sense of achievement and recognition that the team.

Focus also means more collaboration within the team, which can only bring a better solution – creatively built from multiple perspectives.

The intuitive assumption is that the sooner we start something the sooner we finish it…

… the real challenge is how to adjust the behaviour that’s driven by this assumption, which often recurs at every level of the organisation.

.

Tags: Kanban | Lean | Visualising Agile

Friday, 23 November 2012 11:32:04 (GMT Standard Time, UTC+00:00)  #    Comments [0]


# Friday, 09 November 2012

Many turn to contracts as a way to reduce risk. Agile has several alternative, and proven, was to reduce risk. This blog outlines shows how to get from a contract to an ordered list of clear requirements and why that is a better way to define what functionality your customers would value – and then deliver it.

The following recipe can take us from that contractual mind-set to a more palatable Agile one.. by reducing the bitterness of risk.

Step 1: Take one contact (any size will do)

image

… and flatten it out to see its many pages of requirements..

image

 

 

 

 

 

 

Step 2: Go through document identifying the specific requests for functionality and value. Shade the more  important / valuable requirements in a darker shade.

image

Step 3: Boil off the narrative – to be left with the essence of the requirements. Making it easier to read and comprehend. As independent elements in a list they are also easier to keep fresh – with minor clarifications, instead of having to reissue multiple versions of a document.

Risk reduction: the requirements are more easily comprehended in this form – free of the padding narrative, and sometimes semi-deliberate obfuscation, in a document. Stakeholders can get a better idea of what is being delivered – further helped if we use a visualisation technique like story mapping. Hence reducing the risk of misaligned expectations.

Step 4: Further refine requirements for readability – by writing them in simple, non-technical English, from the perspective of a human user and understandable by everyone. – i.e. user stories.

Step 5: Disambiguate* by paring down stories with the use of clear acceptance criteria

(* yes, disambiguate is a word http://dictionary.reference.com/browse/disambiguate)

Step 6: Line up your requirements/stories with the most valuable at the top –defining valuable in whatever way makes sense to your business… it’s not just about money and revenue

Risk reduction: The requirements are clearly prioritised, allowing stakeholders to see and question earlier decisions, thereby reducing the risk of abrupt changes of strategy and direction.

Risk reduction: Developers work down the list from the top, making sure the most important functionality is delivered first. This is very different from when a developer team is given a document of unprioritised features, where they may choose to tackle the riskiest or most intriguing part first.

image

Step 7: Start to deliver the most valuable stories first. Note we say ‘deliver’ not ‘build’ or ‘develop’ or ‘code’. 

Focus on the first shippable slice of functionality – and get it to your customers as fast as possible, to deliver value, learn from feedback and validate your assumptions. This is usually called an MMF (Minimal Marketable Feature) or (MVP) Minimal Viable Product – as advocated in ‘Lean Startup’.

Step 8: Continue to slice the stories down to clarify and ensure you can deliver several (5-15) every iteration.

Alastair Cockburn’s Elephant Carpaccio http://alistair.cockburn.us/Elephant+Carpaccio+exercise is a good recipe for this

Step 9: Slice the stories vertically – so that a user can validate the work and give feedback to the team. Slices should be end-to-end, going through each layer of the cake.

(Excuse the mixing of cake and elephant Carpaccio imagery there.. but you get the point – cake Carpaccio would crumble)

image

Avoid slicing horizontally i.e. spending 4 weeks building out all the database first – because that doesn’t deliver any value or even prove that the functionality works

Step 10: Working down the list from the top the business may decide that they have enough functionality to go live. You may not deliver the lowest value elements – and that’s great. If the dish is good enough to eat – why waste time and money on that last bit of seasoning… (OK, that might be stretching the recipe analogy..)

Risk reduction: Avoid building the low value features which will probably never be used. Save the money – twice - firstly to build it and secondly to maintain it, for ever.

Risk reduction: By properly finishing work as we go along, like a good chef, we aren’t left with a huge amount of un-estimate-able testing (washing up) at the end.

This significant unknown (unknowable!) risk of back-end testing is almost eliminated with good engineering practices. Allowing companies like Flickr to deploy to live many times a day.

image

Note: Whilst you _could_ start with a contracted document – a far better place is a user story workshop, with actual users writing stories onto cards. These are much quicker, more energised and almost certainly generate a better set of requirements to start the project with.

There are many other ways in which Agile reduces risk – this blog has only looked at a few of the upstream contracting elements.

Whilst this recipe won’t guarantee a good experience, hopefully it illustrates how the contractually minded a new perspective on risk management in software development.

.

Tags:

Friday, 09 November 2012 15:32:23 (GMT Standard Time, UTC+00:00)  #    Comments [0]


# Thursday, 01 November 2012

The delivery pipeline is an idea which resonates with our clients. The simplicity of the metaphor and the truths that it reveals make it a compelling message to share with others in the organisation.

The pipeline helps to visualise two key Lean principles; seeing the end-to-end value stream and ‘optimising the whole’ http://bit.ly/ImpLean It illustrates the emerging emphasis on Continuous delivery’ http://bit.ly/CDAgile

The overall organisational aim is to balance the pipeline, as any constrictions generate waste.

The pipeline helps with the key challenge of getting senior stakeholders to prioritise and budget the money, and time, to balance the pipeline end-to-end.

clip_image002

Instinctively, because developers create functionality then, if we want more functionality, we must need more developers. Whilst jarringly simplistic, this is a perspective held by too many senior stakeholders.

However, adding more developers or teams, immediately leads to an imbalance in the capacity of each section. More development teams develop more code, which we then try to force through the lower environments – where it sits, wastefully, in queues or trying to muscle its way past other projects waiting for a limited slot in a integration test environment.

clip_image004

In addition to the waste, the quality plummets due to the stress of trying to get work through this constriction. A situation that is often made worse when these downstream environments are shared with other projects / programmes.

Similar constraints usually exist upstream of the developers. The Product Owner role is new to most businesses and the level of on-going work they need to do is often underestimated. Beyond the responsibility for preparing stories every sprint, collaborating with the team on a daily basis, POs need to actively engage with stakeholders and users.

The diagram below illustrates the actual capacity of the pipeline, which can be quite an attention grabber. It’s troubling for senior stakeholders to realise they are paying for all that under-utilised development capacity.

clip_image006

 

Steps to a balanced pipeline

Step 1: Communicate the need of optimising the system end-to-end

Use the pipeline diagram to communicate the essence of the challenge. These issues are difficult for even IT people to see and understand, even though they live with the stress of these systemic shortcomings on a daily basis. Create a sense of urgency by providing estimates of how much this waste is costing, both in money and in delayed value to customers.

Step 2: Prioritise and target specific interventions at the points where they will have most impact


clip_image008

  • Investment in the environments, automated testing and deployment tools to ensure frequent/continuous movement of software through the environments to live
  • Improve the quality and frequency of collaboration across traditional organisational boundaries, e.g. between developers, service introduction, support and infrastructure.
  • Improve ways of working upstream between Product Owners, stakeholders and the teams. A clear vision with stable portfolio backlog is more likely to create a steady flow of high priority work into the development teams.
  • The teams themselves can improve quality and throughput by adopting Agile/XP technical practices and working more closely with both the product owners and the ‘downstream’ teams.

Step 3: Visualise the waste with a value stream map

Visualisation and quantifying of the waste is key to building a sense of urgency for change and then targeting interventions at the most wasteful sections of the pipeline.

Value stream mapping is the simplest and most effective way to identify where the biggest queues and worst bottlenecks are in the system. The best resources on value stream maps are Mary Poppendieck’s books, including Implementing Lean Software Development’ http://bit.ly/ImpLean

Real-time visualisation of queues is best done with Kanban techniques, which can highlight emerging issues in the system.

Step 4: Get senior stakeholder support and the budget and time

The level of investment needed in environments and tools are often new to a business, which may not be clear on the sizeable returns available. These returns are often in the form of reduced waste, which many senior stakeholders can’t see or evaluate – even though they can sense waste all around.

A senior director estimated that 50% of their software development expense was wasted. How have we ended up accepting, or not able to deal with, this level of waste?

A smoother flowing pipeline also reduces cycle time, meaning that profit-generating functionality is deployed sooner.

Whilst getting the budget is difficult enough, ensuring the work actually gets done, month after month, is even harder, given the prevailing culture in many businesses.

Businesses and managers are rewarded for delivering value to customers. Few businesses reward managers for improving the capability of their system to deliver value.

It’s common to recognise the sacrifice in fixing problems – rather than rewarding those who prevented the problems in the first place.

The ‘non-functional’ work needed to improve system capability is difficult to value and prioritise in any company-wide backlog of work. Those that understand the potential benefits of a balanced pipeline need to articulate that value and be prepared to be held accountable for delivering the promised benefits.

This cultural bias, towards delivery and not improving the means to deliver, continually sabotages attempts to balance the pipeline.

Step 5: Expect resistance and be in it for the long haul

The value stream maps highlight that many constrictions occur at existing boundaries within the organisation. The delay, hand-offs, lack of visibility and accountability at those boundaries are the biggest challenge to creating a balanced pipe.

These changes to what people do, how they work with others, changes to accountability and their daily work will generate resistance.

Specific action is the best way to ignite the change. People on both sides of the boundary typically suffer from its existence and are often keen to improve the situation.

And… it’s not just the pipeline… but the shape of what you put down it..

In addition to working on the pipe itself – the throughput will be significantly improved by looking at the material going through it. This is the subject of a follow-up blog, where we’ll look at batch size, variance and work-in-progress as key variables.

In the meantime, Mary Poppendieck already has some good insights on this at http://bit.ly/LeanPipe

.


Thursday, 01 November 2012 15:31:00 (GMT Standard Time, UTC+00:00)  #    Comments [0]