In Part I, we talked about setting up the postmortem meeting and in Part II, we discussed conducting the postmortem meeting itself. In Part III, we’ll focus on creating a deliverable that can be used to exploit both short-term and long-term opportunities through Action Items and Lessons Learned, respectively. Consider the following in completing your deliverable:
- Respect The Framework: The framework focuses on Pluses or Deltas in each phase of the project or incident. Therefore, only capture those practices that were successful and should be kept (Plus) or that were unsuccessful and should either be changed or discarded (Delta), by phase.
- Action Items or Lessons Learned: Every Plus or Delta captured in the postmortem must result in either an Action Item or a Lesson Learned. Action Items are specific actions that should be taken in the short-term and are assigned ONE owner and a due DATE. Lessons Learned are easy-to-understand guidelines or principles that are understandable on their own and do not require extensive knowledge of the project or incident in order to make sense. More on Action Items and Lessons Learned can be found below.
- Simplicity: Use a simple template and keep the contents pragmatic (I’ve attached a sample template, below). Some additional content, such as names/roles of participants, overview of the project or incident, etc., can be beneficial but the primary contents are Pluses and Deltas by phase and resulting Action Items and Lessons Learned. KEEP IT SIMPLE, creating a deliverable that is READABLE and USEFUL.
If you plan to submit your postmortem document to your business partners as part of project or incident closure, you may wish to include additional content. That said, strive to keep the content simple, direct and to the point unless circumstances DEFINITELY require otherwise.
ACTION ITEMS: Action Items can include, but are not necessarily limited to, any of the following:
- Meetings or Discussions: Meetings to discuss a process change or process development, a potential risk mitigation strategy, etc. In other words, setting up a meeting that might have otherwise have resulted in a problem-solving or sidebar discussion during the postmortem. Do NOT conduct such discussions in the postmortem; conduct them on their own, with the appropriate stakeholders in attendance, and call out an Action Item to make the meeting happen.
- Specific Actions: For example, any code changes, configuration changes, etc. Any specific actions required in the short-term to address any issues that were raised during the postmortem or immediately after project go-live or incident closure.
Action Items have ONE owner, who is accountable for completion. Even if multiple resources will be involved in completing the work, there is only ONE person accountable for ensuring the item is completed and closed by the due date.
Action Items have a Due DATE. The postmortem facilitator should have the Action Item owner set the due date, but if the owner hedges, suggest a due date to them and NAIL THEM DOWN. The due date can always change, but get a commitment during the postmortem and DOCUMENT IT.
LESSONS LEARNED: Lessons Learned should be archived in a single Lessons Learned Master Document (e.g., a simple MS Word document with bullet points). You can categorize Lessons Learned by phase, etc., but be careful; the more you play with the list, the more likely you’ll end up making a task out of maintaining the list. Better to maintain a simple list that can be printed out and reviewed at the start of every project as a reminder of what was learned in past projects and what should be considered in the current project.
Each Lesson Learned should be stated very clearly and concisely and should be understandable ON ITS OWN and on a FIRST READ. Lessons Learned should NOT require detailed knowledge of the project in order to gain context and understanding. Therefore, it’s important to edit each Lesson Learned to make sure they are well written. Have someone who is NOT familiar with the project or incident review the final Lessons Learned for readability before they’re added to the overall Lessons Learned Master Document.
Review the Lessons Learned Master Document after each project or incident to ensure there are no duplicates and also to ensure that, as noted above, each lesson is clear and understandable. Production support/service restoration issues will now allow the luxury of a review of the Master Document at issue identification, so support teams may wish to review the Document on a regular basis as a risk mitigation strategy.
The template below is a baseline document to be stolen/edited it as you see fit. As always, I welcome any comments, questions or thoughts and can be contacted at firstname.lastname@example.org. Good luck!
© 2011, Mark E. Calabrese
In Part I, we talked about setting up the postmortem meeting. The postmortem is part of a project or release close out and is also beneficial (and should be used) at the close of any security or production support incident, etc., as part of the RCA (root cause analysis). Bottom line is that any time a project, incident or initiative is closed, there’s an opportunity to capture any lessons learned and leverage the knowledge going forward.
I’m a big fan of the ‘Plus/Delta’ framework for capturing postmortem findings. In this framework, a Plus is any practice, process or finding that worked well and should be kept or repeated in the future. Conversely, a Delta is any practice, process gap or risk that resulted in a problem and should be changed or fixed. Identify Pluses and Deltas for each phase of the SDLC/Project Life-cycle:
You can use this framework to conduct the meeting, as well. Keep it simple; for each phase, discuss and document all Pluses and Deltas. For incident postmortems/RCAs, you can be flexible on the specific phases (as the SDLC or Project Life-cycle may not be appropriate). For example, incident postmortems can make use the following phases:
- Short-Term Resolution (how was the problem fixed?)
- Long-Term Resolution (how do we make sure the problem doesn’t happen again?)
The point is to go through the incident or project in a logical, chronological fashion, capturing Pluses and Deltas.
Here are some guidelines to consider in facilitating the meeting:
- Who Should Facilitate the Meeting: The facilitator should NOT be a member of the project or incident team. The reason that the facilitator shouldn’t be someone involved in the project or incident is that you want THOSE resources to contribute content, while the facilitator is responsible for guiding the meeting, keeping it productive and acting as…..well….as the facilitator. It’s also best to have a neutral party facilitating the meeting in the case of a not-so-successful project or issue resolution. This avoids anyone thinking that, for example, the developers got off easy because the facilitator was the lead developer on the project. The facilitator should be someone familiar with the team who can keep the meeting controlled and productive. You can facilitate the meeting if necessary, but it’s always better if it’s someone at the same level as those on the postmortem team.
- Pre-Meeting by Functional Teams: The postmortem team should not just “show up” for the meeting and figure it out as they go along. Each functional team (e.g., the Development team, the QA team, the network team, etc.) should meet on their own beforehand to discuss their specific observations. During this pre-meeting, each functional team should use the framework below to document their findings, then publish these findings to the rest of the postmortem team in advance of the meeting. This will help stimulate thought and discussion, resulting in a more productive postmortem. Strongly encourage all participants to review the combined findings beforehand. The facilitator can also combine all the findings into a draft document and send it to participants in advance as well as using the document on the overhead to guide the meeting.
- Documenting: Use an overhead projector and a laptop to avoid any double-work in documenting the findings. If that’s not an option, use a white board or easel and get everything written down and into the document as soon as the meeting is over (and throw all that paper away). Only document findings that will result in an ‘Action Item’ or a very clear ‘Lesson Learned’. For example, documenting that the QA team did “a really great job” feels great but it has no pragmatic benefit going forward. The point is to create a deliverable that is useful beyond the project or incident.
Establish and communicate the ground rules in advance of the meeting and review them as the meeting begins. Make sure you follow the same rules for each postmortem. The facilitator MUST be very clear on the ground rules and tactfully enforce the rules while not quashing discussion (particularly with the ‘No Problem-Solving’ rule; that is the hardest to enforce, I guarantee!). Here are some ground rules to consider:
- No Problem-Solving: The purpose of the postmortem is to capture Action Items and Lessons Learned and NOT to solve all the problems that came up during the project or incident. Problem Solving sessions are Action Items and should be conducted outside of the postmortem and will include all stakeholders in resolving the particular issue or process gap. While you will DEFINITELY have people in the meeting who’ll want to tear into a problem and start kicking around ideas, the facilitator has to stop this and instead, note an Action Item to set up a meeting to tear into the problem and NOT do it during the postmortem (if you ever want the meeting to end, anyway!). Trust me – this is the one rule you’ll be enforcing right up to the end of the meeting, so be very tactful and positive about this – again, you do NOT want to quash discussion and have people shut down/not participate.
- Processes, Not People: When focusing on some of the Deltas, avoid focusing on the actions of specific people. Focus on the process gap and not the person. If someone’s actions caused a problem, this is evidence of a process gap that allowed something bad to happen. If someone falls through a hole in the floor, firing the person who fell through the hole WILL get rid of the “faller” but the hole remains; fix the damn hole. Besides, focusing on the people will lead to bad blood both during and after the meeting. Don’t allow this to happen.
- No Distractions: This one is obvious. The “no sidebars” rule is important and, to my mind, goes without saying. I also recommend a ‘no Blackberries/iPhones’ rule. With the sole exception of someone who has a specific reason, production support issue, etc., that necessitates that they keep their phone near them during the meeting, instead set up a “coral” (i.e., a shoe box or something) to house everyone’s phones and allow a break every 45 minutes / 1 hour for people to retrieve and check their phones. This keeps the meeting focused and is respectful of other people and their time.
- Action Items: Action Items have ONE owner, who is accountable for ensuring the action is completed. It’s true that Action Items usually have many stakeholders/participants, but only ONE person accountable for delivering – period. Also, ‘TBD’ is NOT a due date. Due dates have specific dates, months and years. You can always change a due date but do NOT let anyone walk out of the meeting with an Action Items assigned to more than one person and with no firm due date.
- Follow the Framework: Make sure you follow the framework in order to maintain focus. For example, for a project you’ll start with the ‘Intiating’ Phase and discuss what went well and should be repeated/kept (Plus) and what went badly and needs to be fixed/changed (Delta). The idea is to keep the framework simple and clean and focus on identifying the action items and lessons learned.
Basically, your facilitator ensures that the brainpower in the room produces a solid list of Action Items and Lessons Learned and makes excellent use of everyone’s time. The facilitator should ask for anonymous feedback after the close of the meeting – what could have been done better/different? Another good opportunity to capture some learnings. Also, allowing others to facilitate and avoiding the task yourself is a good way to build leadership skills in your team. If your team is new to this process, you facilitating the first few sessions to establish precedence is worth considering.
In Part III, we’ll discuss the postmortem deliverable and I’ll include a sample.
© 2011, Mark E. Calabrese
Part of closing out any project, outage, service restoration team effort or any initiative whether succesful or otherwise (and it’s even more critical on the “or otherwise”) is the postmortem. Too often, this milestone in the ‘Closing’ phase of a project is either ignored or not fully exploited. I’d like to share some guidelines on conducting a succesful postmortem that net useable results.
While the guidelines below are not meant to be all-inclusive (I like to keep my posts short; no more than 2-3 scrolls as my friend Greg Hubbard at Cenage wisely advised). Use your judgement on what to add, modify or ignore. In Part I, we’ll focus on the meeting itself (timing, participants, etc.):
- Setting Expectations: Make it known either at the start of a project, early in the SRT (service restoration team) effort during an outage or just as a general principle of your management style that you will hold a postmortem after project or issue closure. This will help ensure that people will take note of the good, the bad and other learning opportunities.
- When To Meet: Schedule the postmortem shortly after successful project go-live, or as soon thereafter as is practicable. The key is to schedule the meeting so that doesn’t interfere with any critical post go-live activities but soon enough to capture any learnings while they’re still fresh in everyone’s minds.
- Where to Meet: Pick a conference room that has lots of room and ideally, lots of natural lighting. You want to keep people awake and talking in order to get good input that can be used well after the project is completely closed out. Make sure there are plenty of white boards or easels with tear-away Post-It sheets as well as markers that work. Finally, have some coffee, water, pop (or ‘soda’ as you east coast people insist on calling pop) on hand. I’m not a big fan of food at a meeting, but that’s your call.
- Conference Calls: Ideally, you want people to be there in person but these days that’s not always practical or possible. Do the best you can. In Part II, we’ll discuss how to ensure good engagement and input from those dialing in.
- Who To Invite: Include everyone on the project team, especially business partners who were critical to both the decision-making process as well as in any UAT, JAD sessions, requirement gathering/approvals, etc. Ultimately, you want to include everyone who did the “real work” on the project, as well as any key stakeholders or project influencers. If you find that the list of attendees is becoming unwieldy, try to limit it to those who have the most knowledge of what went right and wrong on the project and also make sure they are people who will SPEAK UP AND GIVE THEIR OPINIONS!
- Timing and Duration: Set aside at LEAST two hours and try to hold the meeting on a Tuesday or Wednesday morning. Mondays tend to be busy and Fridays, people have other things on their minds (or may be working from home). I prefer mornings, as afternoons can be “sleepy time” for some people, so earlier in the week and early in the day is optimal. I like 9AM until 11 or 12.
In Part II, we’ll talk about conducting the postmortem meeting and in Part III, we’ll talk about the deliverables and putting them to work for you, your team and your organizations beyond the project so that you truly CAN leverage lessons learned. Stay tuned.
© 2011, Mark E. Calabrese
I had an interesting coffee meeting today with Gregg Wheeler with Solstice Consulting. Solstice is increasingly becoming known as a solid Agile partner, both in applying the principles to their work and also in helping technology teams apply these principles to their own methodologies. They’re hosting a workshop in April – find more information here.
Gregg and I discussed how Agile is now viewed not as the label to “keep things the same” as was the fad with many technologists in the 1990’s. Now, both technology AND business teams are seeing the value in applying these principles that allow the business-technology partnership to quickly adapt to and exploit business opportunities before the competition. Below are a few frameworks that are worth posting, reading and if applicable at YOUR shop, are worth swiping. All of them are broken down into that most magical of numbers – three:
THREE ROLES ON THE AGILE TEAM
- Product Owner: Plain and simple, this is the business partner who has the business need and decision-making authority.
- Scrum Master: The “project manager” or team lead, whose primary function is to remove obstacles to progress and to coordinate discussions, decisions and where necessary, activities.
- Project Team: The people doing the work.
THREE ARTIFACTS FOR THE AGILE TEAM
- Product Backlog: The prioritized “wish list” of enhancements, features, maintenance items, etc., that are required to meet the businesses needs.
- Sprint Backlog: The next level of granularity, consisting of a breakdown by task with ETC (estimate to complete)
- Burn-down Chart: The breakdown of tasks against the available hours. For example, if a given Agile cycle is two weeks or X hours, the burn-down chart shows how the available time in the cycle will be consumed by the tasks within the Sprint Backlog.
THREE TYPES OF AGILE TEAM MEETINGS
- Sprint Planning: Breaking down the Sprint Backlog into the requirements that will ultimately drive the tasks.
- Daily Scrum: Led by the Scrum Master; these are typically 5-10 minute “stand up meetings” where the Scrum Master asks three key questions: 1.) What did you do YESTERDAY?; 2.) What is planned for TODAY?; and 3.) What OBSTACLES to progress need to be removed (the role of the Scrum Master, as noted above).
- Demo and Retrospective: The team meets with the Product Owner to demo the work to date and to perform a brief postmortem to identify lessons learned and action items with SINGLE owners (a single point of accountability) and with DUE DATES (e.g., not “TBD”)
I think it’s worth repeating the three questions that are at the heart of the Daily Scrum. Again – they are:
- What did the team to yesterday?
- What will the team do today?
- What obstacles must be removed to ensure team success?
I really like the idea of the 5-10 minute “stand up” meeting. I first read about these in Patrick Lencioni’s ‘Death by Meeting’ – a book I highly recommend as it’s full of good info and written in Lencioni’s parable-style that makes for a better read in my opinion.
Agile is a solid set of principles that, when employed by the right type of team, can not only provide exceptional business value to the “business” part of the business-technology partnership, but also creates a fun and rewarding working environment!
© 2011, Mark E. Calabrese
Preserving our sanity is an important part of being a leader. Part of any good ‘Sanity Preservation Program involves understanding and properly embracing the ‘Power of Indifference’. Let me start out by telling you what this is NOT. The Power of Indifference does NOT mean being indifferent to the quality of our work, our team’s work, or our business partner’s and BT partner’s work. Nor does it mean indifference to the consequences of our decisions, our partner’s decisions, our actions, the actions of others…hey, I could go on (and often do) but not now.
There is power, liberation and sanity in being able to “let go” and not let things bother you. It’s only a job – it’s not life. Just because work is part of your life (the part that provides the funding, anyway), doesn’t mean that work IS your life. It’s not. So don’t treat it as such. You owe it to yourself to keep things in perspective and to preserve, protect and defend your sanity.
For most of us, almost nothing we do at work really matters. Our work doesn’t end
war, doesn’t cure cancer and doesn’t make a better world for our children. Some day in fact, all of our work will be thrown away and mocked by people who never even met us and life will go on. Ultimately, there are only two things that matter at work:
- The relationships we build; and
- The deals we broker
Both of these help us grow as people and increase the value we can bring to our teams, each other and our partners. So in the end, work is but the stage upon which we play out the drama of our careers – the value lies in your ability to keep things in perspective and to control events, rather than letting events control you. When you are on the cusp of frustration, anger, or fury of the Russel Crowe or Mel Gibson variety, ask yourself this
question (and you have to ask it exactly like this):
“What difference does it REALLY make?”
Keep this in mind: Always be passionate about work; NEVER be emotional about work.
© 2011, Mark E. Calabrese
I originally read this in a WSJ article back in 2005. The context is initial one-on-ones with your new team. I’ve used these questions with much success and highly recommend them. Make sure you ask them in the order below, as the first few questions will build some trust and ultimately should net you some good answers. Make sure you listen and allow no distractions, but don’t take notes – that will only freak out your new directs. An off-site meeting, like a lunch on you, is one good venue. The questions are:
- What do you want to keep / stay the same?
- What do you want to see changed?
- What do you hope I’ll do?
- What are you afraid I’ll do?
- What questions do you have for me?
As with , any of my other posts, feel free to swipe as you see fit. Good luck!
Apologies to anyone following my blog, wondering why there haven’t been any recent posts. My motherboard died on my Sony Vaio. I took my laptop to Abt (for those of you who don’t live in Chicago, Abt is the electronics toy store for adults – let’s face it, it’s mostly for guys, as we like things that plug in, light up and make noise), as I have an extended service contract with them. At first they told me they’d put a 2-day rush on the part. Then a few days later they told me it would be 7-10 days. THEN they told me it would be a few more days. THEN they told me they’d give me a loaner….then I found out that they had receive the motherboard and installed it…..but that the keyboard and mousepad were also fried and that those parts were on order.
Mind you, this all transpired over two weeks and one day. I’m starting my third week with no laptop, but through the generosity of a good friend, I finally have a loaner. I’ll be calling Abt later today to find out if the parts will REALLY be in on Tuesday and if I’ll get my laptop back on Wednesday. In the meantime, I’ll be using this very nice temporary laptop, which is far, far better than typing up all my emails on my Blackberry using my thumbs.
HAT TIP: To Art Hopkins and Steven Getto for the spell check on ‘hiatus’!