Getting Started With The Clash Retrospective
The Clash Retrospective is a method people use to look back at past sessions and figure out what actually happened versus what everyone remembers happening. I spent years using it in team environments where post-mortems always turned into blame games. The Clash Retrospective strips that away by forcing structure on the recall process. Here is the basic breakdown. You take a completed project or sprint and break it into phases. Then you map what was planned against what actually shipped. The output is usually a simple table with columns for planned items, delivered items, blockers encountered, and lessons learned. Nothing fancy. That is the point.
The Clash Retrospective Format
I recommend starting with this layout: You can build this in any spreadsheet or document tool. Notion, Google Sheets, Excel, even a shared doc. The format matters more than the platform. Step one: gather your data before the meeting. Pull commit history, ticket closures, release notes, chat logs if you keep them. Most people skip this and just show up to talk, which turns The Clash Retrospective into a therapy session instead of a useful review.
Step two: schedule 45 minutes max. Longer than that and attention drops off and people start reinventing problems that were already solved three cycles ago. I set a hard 45 minute limit and stick to it. It forces you to cut the noise. Step three: fill in the Intent and Reality columns together as a group. This is where disagreement shows up naturally. Different people remember different things. Write both versions down instead of arguing over which is correct. The gap column exists for that reason. Step four: pick two or three action items. Not ten. Two or three. You will not follow through on ten items. I learned this the hard way after running a Clash Retrospective that produced a massive list nobody implemented. We missed three consecutive retrospectives after that because the previous one felt pointless.
Get the Full Details
Common Problems And Workarounds
The biggest issue I ran into was when team members inflated their own contribution numbers during the Intent phase. Someone would claim they started work on something weeks earlier than they actually did. This skews the entire comparison. My workaround was to anchor everything to timestamps in the project management tool instead of relying on memory. If it is not in the system, it did not happen for the purposes of the review. Another problem is when people treat The Clash Retrospective as a complaint session. You will hear the same grievances repeat every cycle. When this happens, interrupt and ask the group to convert the complaint into a specific future action. "We have too many meetings" becomes "we will cap synchronous meetings at three per week starting next sprint."
When The Clash Retrospective Does Not Work
This method assumes a baseline of honesty and a willingness to change. If your team culture punishes bad news or rewards spin, The Clash Retrospective will produce garbage output. I have seen it happen in organizations where leadership treated retrospective findings as performance ammunition. The data becomes useless in that environment. In those cases, switching to anonymous input collection helps, but the underlying cultural issue still needs addressing separately. It also does not work well for solo projects. The Clash Retrospective is designed for group dynamics. If you are working alone, a simpler personal review format works better. Track your own intent versus reality without the social layer.
A Real Example From My Work
Last year I ran The Clash Retrospective on a six-week integration project. The data showed we had planned to complete API authentication by week two. We actually shipped it in week four. The gap column revealed that a dependency on a third-party provider caused the delay, but nobody flagged that risk during planning. The action item we wrote was to add a third-party dependency risk assessment to our kickoff checklist. We used that checklist three weeks later on a different project and caught a similar issue before it became a problem. That is the thing about The Clash Retrospective that people miss. The value is not in the document you produce. It is in the specific action items that change future behavior. Without that forward loop, you are just documenting failure instead of preventing it.
Resources
There is no official software for The Clash Retrospective. It is a process, not a product. I have shared templates on GitHub under my username. Search for "clash-retrospective-template" to find a plain CSV version and a Notion version. Both are free. No account required for the CSV.
