Field story 4 min

# The Better Practice Was Already in the Division

Inside one division of a global research and advisory firm, three closely linked teams had each built their own way of taking in and tracking work. The fix was not one team's process imposed on the others. It was finding the best practice that already existed among them and agreeing which few things had to be the same.

The condition

Three teams that shared customers and scope kept their requirements in three different places, so work that crossed between them could never be seen whole.

## Same Customers, Separate Rules

A leader in the data division of a global research and advisory firm asked us to work with three of her teams. One produced and maintained the reporting that turned raw data into insight colleagues could act on. The other two managed data, shared a Scrum Master, and often worked across each other's scope. All three served overlapping internal customers and sometimes depended on each other's output.

They were not, in any meaningful sense, operating as one unit. Each had developed its own way of doing things. One of the Scrum Masters described the result plainly: where requirements lived depended on which team you asked. One team logged everything in its tracking tool and a shared wiki. Another kept notes inside the tickets. A third used a spreadsheet that only two people could open. When something crossed teams, nobody could see the whole of it.

The cause was autonomy applied where the teams could not actually work independently. Each team had built a reasonable system for itself. Together, those systems made cross-team dependencies hard to manage and left leadership without a coherent picture of what the division was doing.

## Two Pipelines, Only One Visible

We proposed two workshops per team. A charter would capture the current state: who the team was, what it worked on, how it operated, and where it thought the problems were. A working agreement would set the future state: how the team wanted to operate and what standards it would hold itself to. The charter surfaces the gaps. The working agreement closes them.

We started with the reporting team. Its charter session opened with a question the team had never been asked together: what is your purpose? Working through the answer, the team realized its internal stakeholders knew what they asked for but not what the team actually built or why it mattered.

The requirements section produced the most revealing conversation. The team described two pipelines. One was the formal request channel. The other ran through chat messages, hallway conversations, and stakeholders who knew someone on the team and had learned that a personal relationship was faster than the official process. Both pipelines carried real work. Only one was visible in any system. When we asked how the team decided between them when they conflicted, nobody had a clean answer. In the closing vote on concerns, the team's top two were conflicting priorities and unclear ownership. Both had been around long enough to feel permanent.

## Commitments the Team Wrote Itself

In the working agreement session, the team turned that picture into commitments. It chose a new name that described what it did, in place of an internal designation that described only where it sat in the organization. It set communication norms for members spread across distant time zones, including response-time expectations, so a message sent at the end of one person's day did not demand an answer from someone who was not yet awake.

It agreed which tool was for which purpose: requirements in one place, project notes in another. Not both, not neither, and not whatever was convenient in the moment. And it wrote separate definitions of ready and done for each major type of work it handled, data capture, dashboard development, and architectural changes, because a single standard had been leaving work ready by the general rule and still missing what that kind of request specifically needed.

## The Answer Was Already in the Division

When we ran the same workshops with the two data management teams, we found their requirements gathering was considerably more structured. Without anyone designing it centrally, they had developed a more disciplined intake practice, with clearer fields, clearer expectations, and clearer handoffs. The reporting team had been struggling with a problem its neighbors had already solved, and neither side knew it.

We brought that back to all three Scrum Masters in one session. Each group had something the other lacked. The goal was not to make everyone do everything the same way. It was to see what was working anywhere in the group and adopt the parts that were better than current practice. One of the Scrum Masters checked the point: they were not standardizing down to the lowest common denominator. They were standardizing up, to a practice that already existed and only needed to be found.

The principle the teams kept was unified, not uniform. A few things had to be consistent: how work came in, what done meant when it went out, and how the teams communicated when work crossed between them. Inside those limits, each team kept the autonomy that made it effective.

## A Shared Language at the Seams

By the end of the engagement, each team had a charter that made its purpose and operating model visible to itself, to the other two teams, and to the division's leadership. Each had a working agreement recording the standards it had chosen. And for the first time, the three teams had a shared way to talk about the work that moved between them.

This work did not end with a headline number. What changed was what could be seen. Requirements had one home in each team. The informal pipeline had a name, so it could be weighed against the formal one instead of quietly winning. And a practice that had been working in one corner of the division was no longer invisible to the team that needed it most.

That is the quieter form of duplication. Not two teams building the same product, but several teams each solving the same operating problem alone while the answer already existed nearby. It ends the same way the louder kind does: by putting the work, and the way the work gets done, where everyone who depends on it can see it.

Teams that depend on each other cannot each keep a private map of the work. Standardize the few things that cross the boundary, and standardize up to the best version already in the room.

Know someone running into this? Send it to them.

[Share on LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=http%3A%2F%2Fagileon.test%3A64380%2Finsights%2Ffield-note-055-one-backlog%3Futm_source%3Dlinkedin%26utm_medium%3Dshare%26utm_campaign%3Dinsight-share%26utm_content%3Dfield-note-055-one-backlog)

Email

### Share With Your Team

Choose who it is for. You send it yourself from your own email; Agileon never writes to your colleagues.

Your Message

Open in Email

Share

[Share on LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=http%3A%2F%2Fagileon.test%3A64380%2Finsights%2Ffield-note-055-one-backlog%3Futm_source%3Dlinkedin%26utm_medium%3Dshare%26utm_campaign%3Dinsight-share%26utm_content%3Dfield-note-055-one-backlog)

Email

Components

[Flow](https://www.agileon.com/insights?component=Flow)[Surface](https://www.agileon.com/insights?component=Surface)

Seen in practice
[The work that fixes this Value Stream Management](https://www.agileon.com/professional-services/value-stream-management)

Keep reading
[Surface The Next Reorg Will Be AI-Driven. Most Will Do It Wrong.](https://www.agileon.com/insights/the-next-reorg-will-be-ai-driven)[Surface By the Time You Feel It, the Cost Is Already In](https://www.agileon.com/insights/field-note-001-demand-already-cost-you)[Surface From 83 Observations to Five Things to Build](https://www.agileon.com/insights/from-eighty-three-observations-to-five-things-to-build)[Flow Surface the Work Before It Surfaces as a Crisis](https://www.agileon.com/insights/field-note-010-before-it-is-a-crisis)
