Analyze Phase Plan: Turn Six Loose Intentions Into an Owned, Dated Plan

The Analyse phase is where a project either finds the real cause or quietly drifts. Teams arrive with energy, a pile of data and a short list of suspicions. What they often lack is a written plan that says who is doing what, by when, and what evidence will exist at the end.

The Analyze Phase Plan tool builds that plan. You list the activities you intend to run. For each one you record an owner, a target date, a status and a note. The tool then calculates how complete the plan is and where the gaps sit. It exposes missing responsibilities, missing dates, missing evidence and delivery risk before investigation begins.

Use it at the start of the Analyse phase, immediately after the Measure tollgate closes. Use it again mid phase when dates slip and ownership quietly shifts. It also works as the backbone of your Analyse tollgate pack, because a reviewer can read the whole plan and its status in one screen.

The tool offers an AI review that explains the calculated result in plain language. The numbers themselves come from your rows.

Everything below is built on the sample data that loads with the tool. The case is fictional. It is used here only to show how the inputs produce the result.

The scenario is a process improvement project with an order entry problem. Baseline work is done. The team has a measured starting point recorded as a baseline DPMO of 12,500. The Analyse phase now needs to move from that baseline to a validated, prioritised set of root causes, and then to a tollgate review with the sponsor.

Six activities carry that journey. They run from early March through to late March. Roles involved are a Green Belt, a Process Analyst, the team as a whole, a Black Belt and a Master Black Belt. One activity is finished. One is under way. Four have not started.

Each row needs four things plus a note. The activity describes the work. The owner names who carries it. The target date commits to a finish. The status says where it stands today. The note holds the evidence or the method.

In the sample plan, row one is Review baseline data and process metrics. Owner is the Green Belt. Target date is 2025-03-01. Status is Completed. The note records Baseline DPMO = 12,500. That single note is what turns a tick box into evidence. Anyone reading the plan knows what the activity produced.

Row two is Conduct value-add / non-value-add analysis. Owner is the Process Analyst, target date 2025-03-05, status In Progress, note Focus on order entry step. Row three is Identify potential root causes (fishbone), owned by the Team, due 2025-03-08, Not Started, with the note Cross-functional workshop.

Rows four to six carry the analytical weight. Validate root causes with data sits with the Black Belt, due 2025-03-15, with the note Use hypothesis testing. Prioritise vital-few root causes sits with the MBB, due 2025-03-20, with the note Pareto + impact matrix. Prepare Analyze tollgate review returns to the Green Belt, due 2025-03-25, with the note Present to sponsor.

The result panel converts those six rows into a small set of figures. There are 6 planned activities. Completion sits at 17 per cent, described as 1 of 6 activities. There is 1 activity in progress. There are 0 blocked or at risk.

The plan definition score is 100 per cent. The tool states the rule behind it directly: it requires an activity, an owner, a target date and a status for every row. All six rows meet that test, so the score is full. The tool also reports no obvious evidence gaps in the recorded rows, because every row carries a note.

Below the headline figures the tool reprints each activity in order with its status badge, its owner and date, and its note. That list is the readable version of the plan. It is the part you paste into a tollgate pack.

Read the two scores as answering different questions. Completion at 17 per cent tells you how far the work has travelled. Plan definition at 100 per cent tells you how well the work is described. They are independent. A plan can be fully defined and barely started, which is exactly what this sample shows.

A result like this one supports a specific set of choices. It does not make them for you, but it narrows the conversation.

First, it supports a go decision on starting investigation. Plan definition is 100 per cent and nothing is blocked or at risk. There is no missing owner to chase and no undated activity to argue about. The team can begin.

Second, it supports a sequencing conversation. Only 1 of 6 activities is complete and only 1 is in progress. Four activities sit Not Started with dates between 2025-03-08 and 2025-03-25. You can ask whether the fishbone workshop on 2025-03-08 has a room, a date and the right cross-functional attendance, because everything after it depends on it.

Third, it supports a load check across roles. The Green Belt owns two activities, the first and the last. The Black Belt and the MBB each own one of the two analytical steps. You can see at a glance whether one person carries too much of the critical path.

The most common mistake is writing an owner as a function rather than a person who agreed. In the sample, row three is owned by Team. That is legitimate for a workshop, but it is also the row most likely to drift, because no single name carries it. If an activity is owned by everyone, check that someone has agreed to convene it.

The second mistake is leaving the note blank. The tool reports evidence gaps based on recorded notes. A plan with full ownership and dates but no notes still passes the definition rule in part and still leaves a reviewer guessing. The sample avoids this by naming the method on every row, from hypothesis testing to Pareto and impact matrix.

The third mistake is stacking dates too tightly. Validation is due 2025-03-15 and prioritisation is due 2025-03-20. Five days separate them. If validation runs late, prioritisation has nowhere to move before the tollgate on 2025-03-25. The tool shows the dates. You have to judge the slack.

The fourth mistake is never updating status. A plan that still reads 17 per cent complete in late March is not a plan, it is a record of intent. Revisit the table weekly.

The Analyze Phase Plan tool checks the structure of your plan. It does not check the quality of your thinking. A plan definition score of 100 per cent means every row has an activity, an owner, a target date and a status. It says nothing about whether those are the right six activities.

It will not tell you whether the baseline figure is correct. The sample records a baseline DPMO of 12,500 in a note. The tool treats that as text. It does not validate the measurement system behind it, nor recalculate anything from it.

It will not judge whether the dates are achievable. It reports 0 blocked or at risk because no row is marked that way. Status is a human input. If an owner knows a workshop will not happen on 2025-03-08 and does not change the status, the result will keep saying zero at risk.

It will not find your root cause. Rows three, four and five describe fishbone work, hypothesis testing and Pareto analysis. Those are separate pieces of work with separate tools. This plan only tracks that they are owned, dated and reported.

Finally, it will not replace the conversation. The strongest use of the tool is to print the result and walk through it with the owners named in the table. The figures open the discussion. The team closes it.