Kanban Template in Excel: How Far a Spreadsheet Board Can Go

A kanban template in Excel is not a workaround. For a lot of teams it is the correct tool, and the reasons are unglamorous: the file already exists, everyone can open it, nothing has to be approved by anyone, and the layout can be changed at four in the afternoon without asking permission. Those four properties beat a feature list more often than software vendors like to admit.

What is worth knowing before committing to one is exactly where the spreadsheet stops being the cheapest option. That boundary is sharper than it looks, and it is not about the number of tasks.

What the board has to do to be a kanban board

Three things, and only three, distinguish a kanban board from a task list.

Work is visible by state. A card sits in exactly one column, and the column is the state. Anyone glancing at it can see where the work is without reading.

Work in progress is limited. Each column has a number, and the number is a rule, not a suggestion. This is the part most templates omit, and it is the part that produces the whole benefit, because the point of the board is to make queues visible rather than to make tasks tidy.

Movement is the update. Work advances by changing column, not by writing a status note. If people are typing sentences about progress, the board is a report.

Any Excel template that supports those three is a real kanban board. Any template that draws coloured boxes and skips the limits is a picture of one.

Two layouts, and the second is the one that survives

The obvious layout puts states across the top as columns and tasks in the cells beneath them. It looks exactly like a physical board, which is why almost every downloadable template uses it.

It has one structural flaw. Moving a task means cutting a cell and pasting it elsewhere, which loses any formatting and any data attached to it, and which two people cannot do at once without one of them overwriting the other. It also gives a task no home for its other attributes. Owner, due date and priority either get crammed into the same cell as the title or live nowhere.

The layout that survives real use puts one task per row with a status column, and renders the board with filters or a pivot.

Layout	Moving a task	Extra fields per task	Holds up with three editors
States as columns, tasks in cells	Cut and paste a cell	Nowhere, or crammed into the title	No
One row per task, status as a field	Change one dropdown value	Any number of columns	Yes

Row per task feels less like a board at first. That is a presentation problem, and presentation is solvable. A cut and paste model is a data problem, and it is not.

Building the row per task version

The whole template is five things, and none of them needs a macro.

Columns. Task, Owner, Status, Priority, Start, Due, Notes. Seven is usually enough. Resist adding more until something is genuinely missed, because every extra column is a field someone has to fill in and will not.

A status dropdown. Select the Status column, then Data, then Data Validation, set Allow to List, and type the states separated by commas in the Source box, for example Backlog,Doing,Review,Done. Keep the in cell dropdown box ticked. This one step does most of the work, because it makes the states a fixed set rather than free text, and free text is what destroys every spreadsheet board eventually. A board with "Doing", "doing", "In progress" and "WIP" as four separate states cannot be counted.

Conditional formatting. Colour the Status cell by value so the sheet is readable without being read. A second rule on the Due column that highlights dates in the past is worth more than any other formatting in the file.

A WIP counter. In a cell above the table, count the rows in each active state. With statuses in column C, =COUNTIF(C:C,"Doing") gives the current count, and putting the agreed limit in the cell next to it makes the breach visible. A conditional format that turns the counter red when it exceeds the limit is the closest a spreadsheet gets to enforcing the rule.

A board view. Filter the Status column, or build a pivot table with Status as the column field and Task as the values, and the row per task data renders as columns without duplicating anything. The filter is less impressive and gets used more.

That is the template. It takes about twenty minutes to build and is more useful than most downloadable ones, because the states match the actual process rather than a generic four step flow.

Three habits that keep the file usable

A template survives or dies on conventions rather than formulas, and three of them do most of the work.

Never insert a blank row as a separator. It looks tidier and it breaks every formula, every filter and every pivot that refers to the range. Grouping is what borders and sorting are for.

Give each task an identifier in the first column, even a plain sequence number. It costs nothing and it means a task can be referred to in chat without quoting its title, which changes. It also makes it obvious when two rows describe the same work.

Move finished work off the active sheet on a schedule, weekly or monthly, into a Done sheet in the same file. A board that accumulates six months of completed rows stops being glanceable, and deleting them loses the arithmetic that was the reason for using a spreadsheet in the first place.

Where the spreadsheet board holds up well

Solo work and pairs. Two people who talk to each other daily need visibility, not coordination features, and a shared file provides visibility at zero cost.

Work with a stable, small set of states. Excel is excellent at fixed structures. A process that has looked the same for two years is a good fit.

Anything that needs arithmetic on the tasks. This is the genuine advantage, and it gets overlooked. Sum the estimated hours by owner, count completed items by week, average the days between start and due. Dedicated board tools mostly require a paid tier or a plugin to do what a spreadsheet does with one formula.

Offline and archival use. The file opens in ten years. A board in a service that shut down does not.

Where it breaks, in the order it happens

Concurrent editing is the first wall. Co-authoring in Excel requires the workbook to be saved in OneDrive, SharePoint or Teams, in .xlsx, .xlsm or .xlsb format, with AutoSave on and edit permission for everyone involved. SharePoint sites not hosted by Microsoft do not support it, and the older Shared Workbook feature is a different, more limited mechanism that Microsoft describes as replaced by co-authoring. When all of those conditions hold, co-authoring works well. When one of them does not, the file gets emailed around and the team now has four versions of the truth.

There is no per card history. A spreadsheet has version history for the file, which answers what the sheet looked like last Tuesday. It does not answer when this particular task moved to Review, or who moved it, or what was said about it at the time. The moment anyone asks why something took three weeks, the sheet has no answer.

Nothing notifies anyone. A task assigned in a spreadsheet is assigned to a cell. Whether the owner knows depends entirely on someone telling them, which means the board only works if there is already a meeting where it is read out.

Discussion has nowhere to go. Comments on a cell exist, and they are not where a team talks, so the conversation about a task happens in chat and the decision never reaches the row. Six weeks later the row says "waiting on client" and nobody remembers what for.

Attachments do not fit. A task usually has a file, a link or a screenshot attached to it. A spreadsheet cell holds a hyperlink, and the hyperlink points at a location that will move.

Phones. Excel on a phone opens the file and is not a tool for changing a status while walking between meetings, which is when statuses actually change.

The arithmetic that is usually skipped

Excel gets treated as free because it is already there. Sometimes that is exactly true, and sometimes it is not.

Excel for the web is free with a Microsoft account, and comes with 5 GB of cloud storage. For a small team's board, that is genuinely a zero cost option, and it supports co-authoring because the file is in OneDrive by definition.

The desktop app is not free. Microsoft 365 Personal is $9.99 a month, Family is $12.99 a month and Premium is $19.99 a month, with Office Home 2024 available as a one time purchase at $179.99. Teams that need the desktop app for other reasons are paying that anyway, so the board rides along for nothing.

The comparison worth running is against the free tiers of board tools rather than against their paid ones. Trello's free plan allows 10 boards per Workspace and up to 10 collaborators. ClickUp's Free Forever plan allows unlimited members with 60 MB of storage. A free plan covering five people and ten boards with every feature present, including Gantt and calendar views over the same cards, sits in the same bracket. Smartsheet, by contrast, offers no permanently free plan, only a 30 day trial, which is worth knowing before starting an evaluation there.

So the choice at small scale is rarely free against paid. It is one free option against another, and the differentiator is which set of limitations matters. A spreadsheet's limitations are history, notification and discussion. A hosted board's limitations are arithmetic and offline access. Teams that need to sum hours by person keep the spreadsheet. Teams that need to know who moved what, and when, do not, and the feature list is the right place to check whether a candidate tool actually records that.

One practical middle path deserves a mention, because it costs nothing to try. Keep the spreadsheet for the arithmetic and move the daily board elsewhere, exporting when a number is needed. Running both for two weeks reveals which one people actually open, and that answer is more reliable than any comparison. What sits above a free plan, if the board turns out to be the one that gets opened, is on the pricing page.

What to change first

Add the two things the template is almost certainly missing: a data validation dropdown on the status column so the states can be counted, and a COUNTIF counter with the agreed limit next to it so a breach is visible. If the board then fails, it will fail on history, notification or discussion rather than on layout, and that is the point at which a hosted board earns its place. Pinateca is one way to see what the same board looks like with those three added.

Q1. Is there a free way to use Excel for a kanban board?

Yes. Excel for the web is free with a Microsoft account and includes 5 GB of cloud storage, which is enough for a board file. It also means the workbook sits in OneDrive, which is one of the conditions co-authoring requires, so a free web setup actually handles simultaneous editing better than a file passed around by email.

Q2. Should the states be columns or a single status field?

A status field with one task per row holds up better. Moving a task becomes a change to one dropdown value rather than a cut and paste, extra fields like owner and due date have somewhere to live, and several people can edit without overwriting each other. The visual board can then be produced with a filter or a pivot table.

Q3. How can a work in progress limit be enforced in a spreadsheet?

It cannot be enforced, only made visible. Counting the active state with a formula such as =COUNTIF(C:C,"Doing"), placing the agreed limit beside it, and adding a conditional format that turns the counter red on a breach gets most of the effect. The rule still depends on people respecting it, which is also true of most physical boards.

Q4. Why do several people editing the same board file cause problems?

Excel co-authoring has conditions: the workbook has to be in OneDrive, SharePoint or Teams, in .xlsx, .xlsm or .xlsb format, with AutoSave on and edit permission for everyone. SharePoint sites not hosted by Microsoft are not supported. When any condition fails, the file tends to get copied and emailed, and multiple versions appear.

Q5. At what point is a spreadsheet board no longer the right tool?

The usual trigger is not task volume. It is the first time someone needs to know when a task changed state and who changed it, because a spreadsheet keeps version history for the file and not for the row. The second trigger is discussion drifting into chat and never making it back to the board.