allan kelly

“the two view-points are always tenable. The one, how can you improve human nature until you have changed the system? The other, what is the use of changing the system before you have improved human nature?”

George Orwell, “Charles Dickens” essay in Shooting and Elephant and Other Essays, Penguin Books

I am sure I am not alone in exhibiting another of Orwellian trait: Double think.

For several years I have been guilty of agreeing with two contradictory points of view:

• The Deming idea that the system governs how people will perform and act
• That no matter what system (e.g. Scrum, XP, Prince2, DSDM, etc.) you use to develop software the motivation and quality of the people is the overriding factor

It wasn’t until John Harvey tweeted about the latter recently that I realised In was double thinking, my positions were not compatible with one another.

I should have noticed the cognitive dissonance sooner. I’ve been known to say, and still believe: “The true test of any software development method (e.g. Kanban, Scrum, etc. etc.) is whether it can help mediocre people perform, i.e. deliver, be more productive”

In that one statement I epitomise the contradiction. I’m hoping that writing this blog will allow me to resolve this!

This isn’t an academic problem: depending on which side of the argument you come down on will determine how you go about improving things. It also explains why some people in the Agile moment emphasis: trust, communication and other inter-personal issues while others emphasis a method, technical practices, etc.

I’m thinking of my experiences:

• I have worked in environments where the system defeated the people: Railtrack privitisation, Sema had an ISO-9000 waterfall process and 120 people. Of course in 120 people some were good and some not so good.
• I have worked in places where (good) people have overcome the system: two years after leaving Railtrack I returned, a much smaller team were in maintenance mode. The same process was in place but we circumvented it, we faked the documentation and so on.
• And I have worked places where everyone was treading water and complying enough with the system

I have seen different things happen within the same organisation, even within what the company regarded as “one team.” I saw Reuters destroy its own processes in an effort to impose a CMMI level 2 (going to 3) process. (Managers were a bit like the American General in Vietnam who supposedly said “we have to destroy this village to save it”. Like the American’s Reuters ultimately lost.)

I have worked places were the concept of “system” would have been laughed it, things might be more accurately characterised as “Chaos.” But even in this chaos there was a system of sorts.

At Dodge Group it was hard to define any system or process, but there was somewhere a process. The company was trying to sell, sell, sell so they could IPO and the software team just had to keep up as best they could.

The good people at Dodge won out for a while. I instigated a simple process we could now recognise as a form of Agile but ultimately the company, the system, killed the success. Partly by hiring bad people, partly by not recognising the good people and partly by letting the system run its course.

In my “Agile will never work in investment banks” posting a while back I made a similar point: Good people in banks can install a good process which can deliver. But in time the bank – the systems, culture and contradictions – will kill the system and dispose of the good people.

Then there is the question of whether “good people” are themselves the result of the system. Can you pick up a “good person” from company A, put them into company B and let them do their good stuff?

An article in the MIT Sloan Management Review a few years back suggested that when “star players” move to a new team they don’t necessarily, or even normally, keep their “star player” performance. (When ‘Stars’ Migrate, Do They Still Perform Like Stars?, Fall 2008 issue.)

So where does all this lead us?

I think there is truth in both the System-first and People-first point of views. Things are not so black-and-white. Sometimes the System-first view is right and sometimes the People-first is right. There are forces at work in different contexts.

Some of these forces are:

1) The strength of the system: the stronger the system the more likely it is to win out.

The strength of the system derives form multiple sources including: the effectiveness of the system, the degree to which it is enforced and the degree to which people see their interests being served by the system.

Reuters CMMI initiative also put a lot of effort into policing. At Dodge the system was weak day-to-day, developers got on with it and they made good stuff.

At Sema Railtrack (first time) a lot of effort was put into policing the system, the system delivered ISO-9000 which helped the company win contracts. As a programmer woking in the system it didn’t seem particularly effective but I guess it was effective at generating money.

2) The System affects the Good People

At Reuters the system encouraged people to leave, I was one. At Dodge the system eventually recruited the wrong people, forced others out and wore down the good people who submitted.

3) Good people can make an poor system work

Primarily through ignoring it, compensating or changing it. Which leads to…

4) Good people can change the system

I’ve seen it happen. In fact you need at least one good person to cause they system to change or it will not.

Again a strong system is harder to change.

5) How “good” are the people?

Yes “good people” can do things change thing but how “good” do you need to be? Is it real a question of having super-stars? Most people aren’t super-stars. Surely we should be more concerned with the mediocre.

And before anyone rushes to say “Good developers are 10 times (or 28 times) better than average” please read Laurent Bossavit book “Leprechauns of Software Development” where he shows the evidence for this claim doesn’t hold up.

I guess these conclusions are going in the Deming direction.

Yet good people can make a bad system work; but if they do not succeed in actually changing the system it will eventually impose itself.

Can I suggest it comes down to a question of time scales: In the short-run good people can overcome a poor system, but in the long-run, unless they change the system for something better the system will overwhelm them.

And the mediocre people? – How about we say this:

Over the short-run, the better your people are the more likely they will overcome a poorer system, but in the long-run the hard it will be for people to sustain the effort if they do not change the system.

A good system will help everyone work the best. It will allow star-performers to shine and mediocre people to contribute and feel good, I’d even like to think a good system will improve the mediocre people.

I’ve written about Conway’s Law before (Return to Conway’s Law (2006) and a Focus Group I ran at EuroPLoP “What do we think of Conway’s Law Now?”)

but I make no apologies for revisiting it again because I still think it hold true.

There are times when I wonder if there is any point to the discipline called “Software Architecture.” Sure design can make a difference in the small but when it comes to the big then, for my money, Conway’s Law is a far more powerful force than the plans, designs and mandates of an enterprise architect.

Conway’s Law predates Agile and Waterfall but it applies to both:

“Any organization that designs a system (defined more broadly here than just information systems) will inevitably produce a design whose structure is a copy of the organization’s communication structure.” Conway, 1968

The original paper “How do committees invent?” is available from Conway’s own website. The original publication was in Datamation, April 1968.

While Conway’s Law can be seen at the macro level, the company level, it is also observed in in the small, at the micro level. The law if often restated in the small as:

“If you have four developers writing a compiler you will get a four pass compiler”

Or

“If you have three developers writing a UI you will get three ways of doing everything (mouse click, menu item, short-cut key)”

For example, if a development team is set up with a SQL database specialist, a JavaScript/CSS developer and a C# developer you they will produce a system with three tiers: a database with stored procedures, a business middle tier and a UI tier. This design will be applied whether it is the best or not. Indeed, the system might not even need a relational database or a graphical interface – a flat file might be quite adequate for data storage and a command line interface perfectly acceptable.

It is a little like a children’s cartoon where the superheroes need to rescue somebody: Superman says “I can use my super strength to lift the ice boulders out of the way”, while Fireman says “I can use my fire breath to melt the ice” and Drillerman says “I use my drill hands to make a rescue tunnel”. Once added to the picture each specialist finds a way to utilise his or her special powers. For a software team this can lead to poor architecture, over complicated designs and competing modules.

In setting up a team architectural decisions are being made even if no architecture diagrams are being drawn. Just deciding to staff the team with four developers rather than three will influence the architecture because work now needs to be found for the fourth team member.

These decisions are made at the time when the least information is known – right at the beginning. They may made by planners – project managers – who have little knowledge of the technologies or by people who will not actually be involved in the development effort, e.g. enterprise architects.

To avoid this my advice is to start a team as small as possible. Try to avoid making architectural decisions in staffing the team by understaffing it. Keeping the team hungry will reduce the possibility of building more than is needed or over architecting it.

There are those who would quickly point out the risk of under architecting a system, not looking to the future and not building a system that can change and grow. But the risks of over architecting are if anything worse. Too much architecture can equally prevent a system changing and growing, and too much architecture leads to more time consuming and expensive code to cut. Then there is the risk of not shipping at all, too long spent producing the “right” design may result in a system too late to be viable.

Second staff the team with people who are more generalist than specialist, people who have more than one skill. Yes have a Java developer on the team but have one who knows a bit of SQL and isn’t scared of a little UI work. In time you might need to add database and UI specialists but delay this until it is clear they are needed.

These suggestions echo Conway’s own conclusion at the end of his paper:

“Ways must be found to reward design managers for keeping their organizations lean and flexible. There is need for a philosophy of system design management which is not based on the assumption that adding manpower simply adds to productivity.”

It is worth remembering that Conway was writing over 20 years before Womack and Jones coined the term Lean to describe Toyota.

Reverse Conway’s Law & Homomorphic force

Since Conway wrote this a lot more systems have been developed. Many of these systems are now referred to as “Legacy Systems.” In some cases these systems force certain structure on the team who maintain the systems and even on the wider company.

For example, take the 3-tier database, business, GUI system from above. Years go by, the original developers leave the company and new staff are brought in. Perhaps some of these have been and gone in the years. The net effect is a system so complex in all three tiers that each requires a specialist. The database is so rich in stored procedures, triggers and constraints that only a SQL expert can understand it. The GUI is crammed full of JavaScript, CSS and not HTML 5 that only someone dedicated to interfaces can keep it all working. And the middle tier isn’t any different either.

Given this situation the company has no option but to staff the team with three experts. Conway’s Law is now working in reverse: the system is imposing structure on the organization.

Again this happens not just at the micro-level but at the macro-level. Entire companies are constrained by the systems they have. Economists might call this path-dependency: you are where you are because of how you got here not because of any current, rational, forces.

Underlying both Conway’s Law and Reverse Conway’s Law is the Homomorphism, or as I prefer to think of it: The Homorphic Force.

This force acts to preserve the structure of system even as the system itself moves from one technology to another, from one platform to another.

Both forms of Conway’s Law and the Homomorphic Force pose a dilemma for any organization. Should they work with the force or try to break it?

I can’t answer this question generically, there is too much context to consider. However I tend towards saying: Work with Conway’s Law, not against it – like woodworking, work with the grain not across it. Be aware of Conway’s Law and Learn to play it to your advantage.

Conway’s Law does contain a get out clause: the system that will be created will be a copy of an existing system, if you can create a new system, a system not pre-loaded with assumptions and well-trodden communication paths then maybe you can create something new and of a better design.

Thus I sometimes view myself as an architect. Now I don’t think I’d accept membership of the Architects-Club even if it was offered, and probably my software design skills are not up to the job but it doesn’t stop me dreaming.

But I architect by (trying) to bring about a good organizational architecture which will in turn bring about a better system architecture via Conway’s Law.

I deliver a lot of Agile training courses and I give a lot of talks about Agile (BCS Bristol tonight). There are some questions that come up again and again which are the result of myths people have come to believe about Agile. Consequently I spend my time debunking these myths again and again.

I’ve been keeping a little list and there are 11 reoccurring myths. There are also two truths which are a bit more difficult for teams and companies to accept.

Agile Myths

1. Agile is new: No, the Agile manifesto was published in 2001, the Scrum Pattern language was works hoped at PLoP 1998, the Episodes pattern language (the forerunner of XP) was workshopped at PLoP 1995, Tom Gilb’s Evo method dates back to 1976 and there are some who trace things back further.
2. Agile means No Documentation: You can have as much documentation as you like in Agile. Documentation is just another deliverable, if it brings you value then schedule it and product it like anything else. Please be aware: documentation is often unread, often fails to communicate, is used as a defensive tool and is typically the second most expensive think on a large software project (after rework).
3. Agile means No Design: No, Agile probably means MORE DESIGN. Design is inherent all the way through development, at every planning meeting and more. Agile does mean the end to big-up-front design which is invalidated five minutes after someone starts coding.
4. Agile means No Planning: No, again, Agile probably has more planning. Again planning is spread out through the whole development exercise rather than at the front and it is the work of everybody rather than one or two anointed individuals.
5. Developers get to do what they like: No, if this is true for you then you are doing it wrong, please call me. Agile needs more discipline from the team and what gets done should be lead from a specific role usually designated the Customer or Product Owner and usually played by a Product Manager or Business Analyst. If developers are doing what they like then there is a failure of in this role.
6. There is a right size for a User Story: There is no right size for a user story. Every team is different, get over it.
7. Work must fit in a Sprint: If you are doing Hard Core Scrum then Yes. If you are playing Agile the way I do (which I now call Xanpan) then No. In fact I advise letting stories span sprints in order to improve flow. You can have stories spanning sprints but we won’t let them continue for ever and we will try and break them down into smaller pieces of work.
8. Scrum and Kanban are sworn enemies: No but the marketing efforts behind each they can get a lot of eyeballs by playing it that way. Xanpan is a Kanban/XP hybrid, and XP isn’t that different to Scrum so there you go.
9. Agile doesn’t work for fixed deadline projects: No, Agile works best in fixed deadline project environments.
10. Agile doesn’t work on Brownfield projects: No, Agile works best in brownfield environments. Granted retrofitting automated unit tests is harder but it is far from insurmountable.
11. Agile doesn’t work on Greenfield projects: No but your first objective is to get yourself to a steady state where you can think like a brownfield project.

To my mind the ideal project to start an Agile initiative with is a brownfield system with a fixed deadline in about 3 to 6 months where development has started but requirements are still unclear.

1. Now two truths about Agile:

1. Agile will not work for us because… (complete this sentence for yourself)
2. Agile is a good idea but … we should wait until we have finished X, got Y to buy in, bought Z and the (new) Pope has given it his blessing

You can always talk yourself out of it or find a good reason for not doing it today.

I’m starting this blog entry on the train returning to London after speaking at Sync Conference in Norwich – and I’m finishing sitting in my local coffee shop on Tuesday. (My conference presentation Agile in 90 minutes is online (PDF) or via Slideshare.)

An awful lot of people at SyncConf worked for Aviva and during the day I spoke to many of them. Aviva are trying to be “Agile” but like every other big company are struggling.

Personally I’m pessimistic about their chances. Someone asked me if I could name a big company which had made the Agile change and I struggled. Yes, many big companies are trying Agile, and yes there are many successful teams in these companies. But has an entire company made the entire change? I don’t know of any.

All the banks are trying it, and all are failing in similar ways (see Reverse-Tolstoy and Why Agile will Never work in Investment Banking). Many other big corporations are trying and most of what I here is not positive.

During the day I did build up a picture of much that is happening inside Aviva and I did give some advice in conversations. Now I’ve thought about it there is some more, bullet points below.

(I can say this because Aviva are not my client – OK, I did do a little work with them years about but it was insignificant. If Aviva were my client I wouldn’t be writing this blog, I don’t speak openly like this about clients. I try to avoid talking about active clients at all in this blog. Nor am I in the habit of giving unsolicited free advice to clients.)

Advice for Top

• There are no Silver Bullets: Conway’s Law, Reverse-Conway’s Law, Brooks Law, Goodhart’s Law and dis-economies of scale constrain what can be done.
• Take Software Craftsmanship seriously: Start and apprenticeship programme – speak to Jason Gorman about how the BBC do it
• Let a thousand flowers bloom: dump any attempts at standardised working and let people experiment, let them have fun as they learn and change. In a few years time when you’ve made lots of progress start to mine out your own “Good practices” and sharing them amount teams.
• (That also means be prepared to work with multiple difference consultants, coaches and advisors. Accept they will give you different advice. Yes engage with some big players but also work with the small players.)
• Embrace dis-economies of scale – you don’t have time to be involved with the details so delegate
• Get with Quality: the Quanlity v. Time trade-off is the reverse of what you think; higher quality (i.e. reduced bug count) makes for shorter scheduled, shorter schedules makes for lower costs, together this make for happier customers and more responsive business
• Get clear about what you want: if it is cutting costs say so, you can use some of the Agile toolkit for this but don’t expect to use it all or achieve Agility. There is nothing wrong with not being Agile if you have a better alternative.
• Pay the SyncConf Norwich team to re-run the conference again (with a few tweaks) inside the company. (You’ll have to pay the speakers but this will be a drop in your ocean of your IT budget.)
• Go further and let give employees attending seed money, let them spend it to sign up for courses and bring in some of the consultants they liked into the company. Yes this is radical, no apologies.
• If you want to be Agile then accept that it won’t be the lowest costs, it should result in the greatest value overall so focus on value not costs
• Your business and IT operations need to be more joined up. Maybe you want to dismantle your IT group, at least embed it in the business.
• Your annual planning cycle needs to relax, you need to learn to start small pieces of work without having every T crossed an i dotted.
• In the long run you want to look to Beyond Budgetting.
• Don’t worry about your CMMI or ISO or whatever accreditation, my clients have FDA and ISO-13485 approval, if they can do it so can you.
• In the meantime give up on strategic planning: that idea died in the paddy fields of Vietnam thanks to Robert MacNamara. Go read Henry Mintzberg “Rise and Fall of Strategic Planning”.

Advice for the Middle

• Visualise, visualise, visualise: boards, graphs, everything!
• You have to lead the top, they don’t know how to spell TDD so educate them
• Involve “the business”, their representatives in everything
• Get with the Quality Agenda: ditto the above, embrace and push quality to eliminate bugs. Give your team support, send them on training, hire consultants to come and coach, buy them books.
• Architects must also code
• Celebrate every little success; make sure the whole company knows about a realise
• Spend on travel: Indian outsourcing doesn’t work because you paid TCS, you actually have to go there and see, they have to come here and see, you need to talk, talk, talk. Skype, VOIP, everything
• Don’t spend money on expensive test tools
• Manage teams as a whole: not a Test Team and a Test Manager there, and a Dev Team and a Dev Manager there, and a BA…. the whole team wins or looses together.
• There is no long term future for Test Managers, you options are: a) focus on Test and work to improve the Testing Process, specifically Automation; b) focus on management and manage the whole development team; c) disentangle Test from Quality and become get Quality throughout the process
• Form self-support book study groups: spend an hour, meet for lunch every other week, agree a book, agree a chapter in advance, someone volunteer to summarise it. Open each meeting by going once around the table and having someone say something positive (“Tell me something that made you smile today?” “Tell me something you learned this week?”).
• Good luck: you have the hardest task, and its lonely. Some of you will need to lay down your jobs to make this happen

Advice for the Workers

• You control the means of product, Agile is in your hands, Make it So. Your greatest weapon is Quality, you have the power to improve it today.
• Embrace Test Driven Development, Refactoring (i.e. emergent design.) Dump big-up-front design, get coding on day-1 and work out the design later.
• You have to lead the middle and they will lead the top. Neither of these two groups knows as much about your work as you do. Lead by showing them what works. Which means…
• You have to do it, make those improvements. Developers start learning TDD, spend your own money if you need to. Testers: download the open source test automation tools and enrol developers to help you get them working.
• Testers: Automation, Automation, Automation – get with the Open Source tools – Selenium, FIT/Fitnesse, Cucumber, etc. etc.
• Form your own study groups
• Make sure retrospectives happen and action items are actioned

Most of this advice is applicable to any large company. I really don’t know that much about the inside of your company, I’m assuming. Nor is this list really that different from the one I published on InfoQ last year – 10 things for making your Agile adoption successful.

Finally, Aviva, you don’t have to take any of my advice but a) I guess your paid advisors are saying similar things and b) if you don’t change new competitors will kill you; you are big so it will take time but it is never nice to watch a big animal in pain.

A question that comes up again and again from teams is “What is the right size for a User Story?” – a day? a week? three weeks? – 1 point? 5 points? 10 points? 37? 100? 5 words? 20 words?

Short answer: there isn’t a universal right size for a User Story, it depends on your team, their skills and their level of domain knowledge.

Long answer: for some teams a User Story is Big, several days or weeks of work. On other teams they are small – maybe a day’s work. Both are possible, both are the right answer. Although, as a general rule smaller is better.

For me there are criteria that a User Story should meet:

• It should be small enough for the technical team to understand it and create in in a short time period
• It is small enough to move on the board sometime soon
• It should be big enough to represent business value in its own right – it might build on something that has been done before (e.g. a lower fidelity version of the same story, the new one increasing fidelity)
• It is big enough to be deliverable in its own right – you might not want to do so but if you needed to you could

In an ideal environment: the story would be developed in a few days, would be released on its own right and the business value measured directly. (For those who know about Minimally Marketable Features (MMFs), also sometimes called Business Value Increments (BVI) or Quantum of Value (QoV), I casually consider a User Story to be a MMF.)

When a team are dedicated to a particular product, and have worked on it for several years, and have learned about the domain – as I would expect in for in-house development operations – I expect to see larger stories. Conversely, when a team don’t know the domain – as is common with outsourced development – I expect to see small stories, and more knowledge pull from the client.

Taking a step back, a few basics:

• A User Story is not in and of itself a requirement; I like to call them “Tokens for Work to be Done”, others like to say they are “A placeholder for a conversation” both are true
• Don’t feel that you must use User Stories: if the format doesn’t fit your need then don’t use it! You can still have a “token for work to be done” and a “placeholder for a conversation” using any words you like, they don’t have to be “As a… I can … So that…”
• The User Story format does capture three very important elements: Who, What, Why so they make a good starting point and are easy to understand but if they don’t work for a particular need then don’t force it into the format

Small stories are better – they can be developed more quickly, delivered more quickly and return value more quickly. But there comes a point were stories become too small to return any business value at which point they, well, loose value. Plus, when you have a lot of small stories they become a headache to manage.

One common problem I see is teams who create really small stories in order to get them to fit within one sprint. These means large stories don’t get scheduled or teams split large stories into many small ones which have no business value themselves.

People say this is a Scrum thing, I’m not sure. Scrum doesn’t mandate User Stories, the story format came along after Scrum. True the format is widely used they has become part of Common Scrum.

What Hard Core Scrum does say is that all the work the team commit to should be done by the end of the Sprint. Which would imply a story has to be small enough to fit I the Sprint. The problem then comes: what else do you put in the Sprint? If one story is going to take more than half the Sprint you need one or more stories to use the rest of it up. Plus you hit the commitment problem – developers have an incentive to under commit and the business/custom has an incentive to over demand.

My solution – one which I’m baking into my description of Xanpan – is:

• Stories are split into Tasks during the planning meeting
• Tasks do not follow any particular format, nor do tasks have any business value
• Tasks are for the team – the developers, testers, analysts and anyone else
• Tasks should be small, a day or two at most.
• A story is not done until all the tasks are done. Stories can therefore flow across iterations, the chances of tasks flowing across iterations is low (because they are small) but they can and do.

However – and this is the key point – in counting the points (velocity) of a team only the completed Tasks are counted. Partially done tasks are not. The only states that matter are Done and Not Done.

As I’ve written before, the breakdown of stories to tasks fills (at least) three purposes:

• It is superficially an estimation exercise
• It is also a requirements elicitation process (I like the Requirements Engineer, BA, Product Owner/Manager to be present)
• It is a design activity.

You know a task is too big because it is difficult to finish. If it is difficult to finish it needed more thought up front, to define and understand it, to bound it, to understand it. (See the mini-series on Story Points earlier this year, and specifically the Story Breakdown entry.)

So there you have it: there is no right size of a User Story.

That said, they may still be “too big” or “too small”. But what constitutes right it dependent on you, your team and how you play the game.

Something I’m very conscious of is that I don’t hear much about the downside of Dialogue Sheet Retrospectives. As much as I’d like to believe that this is because there is no downside I can’t honestly say I think that is the case.

There must be downsides.

I have had several suggestions for improvement and some of these have been incorporated into some sheets. There are also a couple of pending improvements which I’ve yet to action. Because there are several sheets and because not all of them are for retrospectives some improvements get incorporated into some sheets but not others. In part I’m happy with this because it increases the differences between sheets.

At XP Day this week I had several conversations about the sheets and I now think its worth listing some reasons people choose not to use the sheets or negatives that emerge. Some of these I’ve known about for a while and may have mentioned before, some are new.

In no particular order, and in my words summarise what I’ve heard….

“The team looked at the sheet and don’t see how they would be an improvement on the current retrospective format”

This reason has been given by several people/teams and while I respect it as a team’s decision to make I can’t help feeling that it pre-judges the sheets. Surely the “Agile” way of doing it would be to try a sheet and see what happens. After all, what have you got to loose? One or two hours, one retrospective.

“We need a facilitator”

Almost a continuation of the previous one another reoccurring comment is from people who can’t imagine not having a facilitator. Again I would suggest try it and see.

I have observed many retrospective dialogue sheet sessions and I find them very interesting. If a facilitator wants a role I have several suggestions:

• Take off the facilitator hat and join the retrospective, you will get insights for the next time you facilitator
• Observe silently from the sidelines: you will learn things just from observing
• Use the sheets from one retrospective as input to the next, go back and pick up items, delve deeper into issues
• Use the sheets as first exercise in a longer retrospective; do a sheet then go back over issues that arise

You might not need a facilitator but that does not mean you must not have one.

“Facilitator would go deeper”

I am convinced there are issues and times where having a facilitator will produce a better outcome. For example, some issues benefit from talking about in depth. That is why I always thing the sheets are one tool amount many. However I don’t think you always need to go deep. As with the first two items in this list i encourage you to try it and see.

Cross training retrospective facilitators

This might be a rare reason for not using the sheets but is quite valid. One company wanted to train a number of people to under take retrospectives so each retrospective was used as a training exercise for a facilitator.

“We have a company mandated retrospective format and we cannot change it”

This has to be the worst reason for not trying any new technique in a retrospective. If a team cannot change the retrospective format what chance have they got of acting on any findings from the retrospective? Actually I think this suggests a bigger problem.

Mechanical sheet filling

I only have a second hand account of this and I’d like to know more. It appears a team at a large media company in Salford do not engage in conversation in doing the sheet. It appears filling in the sheet has become something of a chore and they simply “do it”.

With out knowing more about this team (anyone know them?) it is hard to say what is going on. It sound like the sheet has been imposed on the team and they are completing it like another piece of administrative paperwork.

Fear of team dynamics

Some people are worried that personality issues on a team will disrupt the dialogue sheet process – conflicts, arguments, etc. I think this is a valid concern. Although I then immediately think of two things.

First: this is an example of a bigger issue – if these people can’t spend an hour together on a collaborative exercise then working together must be very difficult.

Second could this type of exercise actually help with those dynamics? I would guess the answer is sometimes Yes and sometimes No.

Bored by repetition

There are a few teams that have tried the sheets used the them regularly and then they have fallen into disuse. I suspect this is true of any retrospective format that is used too much. The trick is to find whats to keep the retrospectives fresh and keep them generating new insights.

If you know of any downsides or problems with dialogue sheet retrospective then please let me know. I might be able to correct it in a sheet, if I can’t then we can at least understand their use better.

I have started to write up a more useful description of Xanpan. In doing so two things have happened. First I’ve realised that I have an awful lot I would like to say about it, I’m terrified it will become another book. Second, I’m becoming more and more aware of how Xanpan differs from Scrum, XP and Kanban.

As a result I expect this blog will get a little quieter. But before the end of the year I’d like to get a few entries finished and published which have been languishing for a while.

So without further ado here are 10 tips for teams and those who manage, administer or simply organise teams. Of course, if you are a self-managing team you should all read this list.

1. 1. Don’t load overload teams – consider 76% utilisation about the max
   2. Keep teams together – bring the work to the teams, don’t form teams around pieces of work and then break them up when it is finished (Corporate Psychopathy)
   3. Minimise disruption to the team, if the team needs to be regularly interruptible and responsive then set processes and expectations accordingly
   4. The more self-standing, free of dependencies, the team the more effective they will be
   5. The team should own the process and methods of working as well as the things they work on. By “own” I mean: they are allowed to change the process and methods.
   6. Make the boundaries and responsibilities of the team clear
   7. The team needs a full skills set to fill 4,5 and 6; and those with the skills need to be on the team full time
   8. Aim for an MVT (Minimally Viable Team) so put slightly fewer people and skills on than you initially think, expand the team as they show success but always keep them minimal
   9. Aim to recruit slowly and regularly, avoid foie gras recruitment, i.e. stuffing the team full of new people in the hope of increasing quantity in the near future (you won’t, you’ll cripple yourself in the short term)
   10. Give the team as much authority and power as possible to do their work, the more decentralised the organisation the more effective this will be

</ol style=”text-align: left”>Looking at that list two things strike me. First is how many of my older blog entries are linked. Second about three of those items are things not to do. Simply not doing bad/mad/stupid things can deliver benefits.

Since I published Scrum, Scrum and Scrum on this blog last week the piece has received a lot of attention and I’ve had a lot of feedback: comments on the blog, comments on Twitter and comments to me in person at Oredev last week.

On the whole these comments have been in agreement. My impression is that people find the posting a useful distinction between Synonym-Scrum, Hard-Core-Scrum (or ScrumTM as I’ve also called it) and Scrum-Lite (or Common-Scrum as it might better be called). In this blog entry I’d like to clear a few things up and reply to one particular comment from Kurt Häusler which I think deserves more attention.

Just to be clear: I’m not attacking Scrum here. I’m discussing difference in how it is perceived, what people call Scrum, and where I think “Scrum” is inconsistent (OK, that might be criticism.)

First terminology.

In using the term ScrumTM I am being sarcastic. Nobody owns a trade-mark on Scrum, anyone can use the term Scrum and the ideas in it. So lets stick with Hard-Core Scrum.

Scrum-Lite also seems to upset people, this name was meant to designate a form of Scrum which is not Hard-Core-Scrum. Lets use the term Common-Scrum because this is the way I most typically see the Scrum ideas implemented.

Next the issue of Project Managers, and to a lesser degree Architects and other specific roles on Scrum. Kurt wrote:

‘Anyway, anyone saying project managers and architects are not allowed in Scrum is simply wrong. Are they really saying that though? And are people saying that really representing a different meaning of the term “Scrum”?’

Taking it in reverse order: I am the one who says there are two-forms of Scrum. I think this makes a useful distinction which helps me, and I think others, understand the world we live in.

Kurt’s first question is the more interesting one, if the answer was “No”, if the world did have a homogenous view of what Scrum is the second question, and the distinction would be unnecessary.

So Kurt, the answer to the first question is a most definite Yes. I think the point is well made by Christin Gorman in her 10 minute video on Vimeo from the Roots conference – “Putting an architect in a Scrum team is like putting mayonnaise in a cake.”

Let me say straight away I’m not picking on Christin here, I actually agree with much of what she says and she is not alone in this view. There are other – better known – names who say the same thing:

• In The Scrum Primer Deemer, Benefield and Larman say “there is no role of project manager in Scrum.” These words are in version 1.1, there is a Version 2.0 on the Scrum Foundation website which adds Bas Vodde to the author list and contains the same words. I haven’t read v2 in detail so maybe in contains more advice for project managers.
• The Scrum Handbook on Jeff Sutherland’s website contains the same words – in fact the biggest difference to The Scrum Primer seems to be the author list.
• I personally heard, several years ago, Craig Larman answer the question “What does a project manager do in Scrum?” by saying “If you have a project manager in a Scrum Team you are not doing Scrum.”

The important point is: Christin understands Scrum to mean “No project managers” and Kurt understands Scrum to allow project managers.

I sympathise deeply Christin’s point at the end of the video, if I may paraphrase: “lets save Scrum to mean what it was meant to mean”. I just happen to think this particular cat is out of the bag. Rather than trying to “save” or “reclaim” the name “Scrum” I prefer to differentiate.

Interestingly, as I have commented before (“When did Scrum start loving project managers?”) Ken Schwaber hasn’t always shared this point of view:

• In Agile Software Development with Scrum Schwaber says “The Scrum Master is a new management role introduced by Scrum” and a little later “The team leader, project leader or project manager often assume the Scrum Master role.”
• In Agile Project Management with Scrum Schwaber says “The Scrum Master fills the position normally occupied by the project manager. I’ve taken the liberty of redefining the role.”

Now perhaps I’m guilty of not checking my facts. I haven’t e-mailed any of these people to ask where they stand on Project Managers and perhaps I should have. (If any of you are reading this please feel free to comment.)

However, I don’t believe a definitive statement would make much difference. What if I had a statement from one of these people saying “Project Managers are in” (or out) ? There would still be people out there interpreting Scrum the other way. It is too late to change this now.

Personally I believe that overtime those who define what Scrum is – those named here plus a wider community – have changed their position on Project Managers. I suspect that in the beginning there was no anti-Project Manager bias. Then is crept in, and now it is less, or perhaps gone altogether.

The points I want to make are:

• there is are different understandings of what Scrum is
• I, and I think others people, find it useful to make a distinction

I really don’t care if you want to say “Common Scrum is not Scrum because…” or “Hard-Core Scrum because…” is not Scrum, because I’m not describing what should be, or not be. What I am describing here is how I find the world.

Comments please?

I’ve come to believe there are three different meanings of the term “Scrum” – well, three inside the software development community at least, if we consider sport we can probably add a fourth.

Even if nobody else does I differentiate between:

Scrum as a synonym of Agile

“Scrum” means “Agile” like “Hoover” means “Vaccum Cleaner” in English English, or like “QTips” means “Cotton Bugs” in American English, or increasingly “Google” means “Internet Search” whether you are using Google, Bing, Yahoo or some other search engine.

This is the way I believe a lot of software people use the term Scrum. Although Agile is more than Scrum alone often (mostly?) when people say Scrum they mean some a general form of Agile, something with iterations, stand-up meetings, perhaps User Stories, perhaps some technical practices, etc. etc.

Hard Core Scrum, ScrumTM

This is Scrum without Project Manager or Architects, maybe even without Testers. This is Scrum of Commitment and Abnormal Termination of Sprints.

I once heard Craig Larman say “If you have a project manager on a Scrum team you aren’t doing Scrum.” Only last week Christin Gorman Tweeted saying: “If you want to use managers and architects, you won’t be doing scrum. And that’s ok.”

I’ve written before about how this hard-core Scrum might actually conflict with Scrum (When did Scrum start loving Project Managers) and XP (Two ways to fill an iteration).

This view sees Scrum as a kind of Communist Manifesto but with Managers as the Bourgeoisie. It would be funny if this weren’t so serious. I believe that if you take this interpretation you will soon run into opposition from a layer of managers. The paradox is that Scrum is as popular as it is because those manager don’t like Extreme Programming and saw Scrum as the management friendly alternative (The Three Advantages of Scrum).

Common-Scrum or Scrum-Lite

This is Scrum but without the anti-manager zeal. Of course for many people this strips Scrum of its primary asset – and they might be right. It might be that if you neuter Scrum you get a less effective result.

In this form of Scrum Project Managers still exist – assuming you had them to start with, a lot of places don’t. However they have been renamed Scrum Masters – see my example from 2008.

In Scrum-lite the word commitment really means “agreement” because the team are probably using velocity to judge how much work they can do. Architects and other ranks, sorry, roles, still exists. And companies still kill teams – corporate psychiatry is rampant.

In other words: it is full of compromises. It works for some companies, it doesn’t work for others and ScrumTM purists will hate it.

Anyway, I find it useful to distinguish between these three uses of the term “Scrum”. If you know any more please add them to the list.