Skip to main content

Data refreshed daily · DR by Ahrefs

Sign in

Building a Launch Dashboard in a Spreadsheet

Dashboards · 11 min read ·

You do not need a data team. A single spreadsheet with a few tiles, a clean data tab and weekly habits can run a launch. Structure, formulas and pitfalls.

Illustration: A light grey dashboard with six KPI tiles at the top, a row of small charts beneath and a tab strip at the bottom labelled Dashboard, Data, Notes, in the teal, amber and slate chart palette

Before dashboards had names, people kept ledgers. A page for each week, a few columns, one running total. The format survived because it worked: you could see at a glance where you stood.

A spreadsheet is the modern ledger, and for a single launch it is usually enough. It costs little, runs anywhere and bends to your needs. This guide shows how to build a tidy launch dashboard in one: what to put on it, how to structure the data, which simple calculations earn their place and the habits that keep it honest.

Start with the questions

A dashboard is an answer to questions. Write them before you open the sheet.

For a launch, typical questions are:

  1. Where are visitors coming from?
  2. Are they doing what we hoped?
  3. Are they coming back?
  4. Is the product link profile growing?
  5. What are people saying?
  6. Is anything broken?

Each question becomes a tile. If a number does not help answer a question you care about, leave it off.

Choose five to eight tiles

A launch dashboard that fits on one screen is read. One that needs scrolling is not. A good set:

Visits. Total visits this week, with the change from last week.

Visits by source. A small bar of the top five sources.

Actions. The count of the one action that matters: sign-ups, trial starts, enquiries.

Conversion. Actions divided by visits, shown as a percentage but backed by the counts.

Returning visitors. Share of visits from people who have been before.

Link profile. Your current Domain Rating with its signed change and check date, plus the count of referring domains.

Feedback. The number of replies and questions this week, and a single line on the main theme.

Issues. The count of open problems.

Eight tiles is plenty. Many launches do well with five.

Structure the sheet in three tabs

A spreadsheet goes wrong when everything is in one place. Use three tabs, each with one job.

Tab one: Data

This tab holds the raw numbers, and nothing else.

  • One row per period, usually per day in launch week and per week afterwards.
  • One column per measure, with a clear name in the top row.
  • A date column in a consistent format, such as year-month-day, so it sorts correctly.
  • Source columns if you track several: visits from each tagged source.
  • No merged cells, no blank rows, no colour as data.

You type or paste in this tab. You never calculate here.

Tab two: Calculations

This tab reads from Data and works out what you need.

  • Weekly totals.
  • Changes from the previous period.
  • Conversion rates.
  • Averages over a rolling window.
  • Any ratios.

Keep each calculation in its own labelled cell. When someone asks "how did you get this number?" you can point to a cell.

Tab three: Dashboard

This tab shows tiles and charts, reading from Calculations. Nothing is typed here.

Keeping the tabs apart means you can change a chart without touching a number, and fix a number without breaking a chart.

The calculations you need

Spreadsheets can do clever things. For a launch dashboard, a handful of simple ones suffice.

Change since last period. This week minus last week, as a number, with a sign. Show plus or minus, not only colour.

Percentage change, carefully. When the base is small, a large percentage means little. A rise from two to four is a hundred per cent and still two people. Prefer counts for small numbers.

Conversion rate. Actions divided by visits. Show the counts beside it.

Rolling average. The average of the last four weeks. A moving average replaces each value with the mean of itself and its neighbours, which smooths short wobbles and reveals direction.

Median instead of mean where a few large values would distort. The median is the middle value when numbers are sorted, and it resists outliers. It is helpful for something like time to first action, where a few very slow users inflate the average.

Share of total. A source's visits divided by total visits.

Write down the formula you use for each, in a notes tab, so it stays consistent.

Tagging, so the data has sources

A dashboard is only as good as the data in it. Tag your links so that visits are grouped by source and medium in your analytics tool. Google's help pages explain how campaign parameters are read. Use a short list of names and stick to it, for example the name of the place in lower case for source and a fixed word for medium.

At the end of each period, copy the totals by source into the Data tab. If your analytics tool can export a table, paste it as values.

Add context with a Notes tab

Numbers need stories. A fourth tab, Notes, holds a dated list of what you did and what happened.

  • The day you posted in a community.
  • The day a listing went live.
  • The day a bug was fixed.
  • A mention you noticed.
  • A change to the page.

When a number jumps, the notes explain it. When you review the launch later, they are invaluable.

Charts that help

Choose charts for what they show, not how they look.

Line charts for trends over time. Use a consistent time axis and label it.

Bar charts for comparing sources or categories. Start the axis at zero.

A single big number for a tile, with its change.

Small tables for exact figures.

Avoid pie charts with many slices, three-dimensional effects, dual axes with different scales and decorative backgrounds. A calm, chart-first layout uses a restrained palette, with colour reserved for meaning.

Label every chart with a title that says what it shows and the period it covers. Add a note on the data source and last update.

Make it accessible and readable

  • Do not rely on colour alone. Use arrows, signs and labels.
  • Check contrast of text on coloured tiles.
  • Use clear fonts and sizes.
  • Add short text descriptions beside charts so the main message survives if the chart is not visible.

If you share the dashboard with others, check that it works on a phone.

Habits that keep it honest

A fixed update time

Choose one time each day in launch week and one each week afterwards. Update then. A dashboard that is updated when you remember will be wrong when it matters.

One person owns the numbers

If several people edit, agree who is responsible for the Data tab. Others read.

Check for obvious errors

After updating, scan for numbers that look wrong: zero where there should be something, a spike that matches no event, a total that does not equal the sum of its parts. Fix before you share.

Keep a version

Save a copy at the end of each week with the date. If a number is ever questioned, you can show what you saw then.

Resist the urge to add

Every few days you will think of a new metric. Add it only if you can name the question it answers and the action it might prompt. Otherwise write it in a "maybe" list.

Review the questions monthly

Do your tiles still answer your questions? Retire any that do not.

A starter layout in words

At the top, a row of five tiles: visits, actions, conversion, returning visitors and DR with its change. Below, on the left, a line chart of weekly visits with a four-week average. In the middle, a bar chart of visits by source. On the right, a small table of the top comments and questions of the week. At the bottom, a strip of notes: the three most important things that happened. That is the whole thing. A glance gives you the state, and a second glance gives you the reasons.

Privacy and data care

If your dashboard includes personal data, such as names or email addresses, keep it out. Counts and aggregates are enough. Do not paste user-level exports into shared sheets. Limit who can view the file, use strong passwords and two-step sign-in on the account that owns it, and delete old copies you no longer need.

When to outgrow it

A spreadsheet starts to strain when:

  • You manage many products or sites.
  • Several people update data at once.
  • You need data refreshed automatically many times a day.
  • The file becomes slow or fragile.
  • You need access control that spreadsheets cannot offer.

Then it may be time for a dedicated tool or a small database. Until then, the ledger works. If you want to automate parts of it, this platform provides a free API and MCP server for reading listing data, described on the developers page, which can fill some of your Data tab without typing.

Common mistakes

  • Too many tiles.
  • Calculations in the Data tab.
  • Merged cells and blank rows.
  • Inconsistent dates.
  • Colour as the only signal.
  • No notes on what happened.
  • Updating irregularly.
  • Chasing percentages on tiny numbers.
  • Putting personal data in a shared sheet.

On this site

The launches and leaderboards pages show how a dense set of numbers can be laid out clearly, with signs and arrows as well as colour, and the developers page describes the free API and MCP server. The submit page starts a listing, and the blog has further guides in this series.

The short version

Write your questions first, choose five to eight tiles and build three tabs: raw data, calculations and dashboard, plus a notes tab. Use simple, documented calculations, tag your links, choose charts that start at zero and label themselves, never rely on colour alone and update at a fixed time. A tidy spreadsheet is a ledger with a view, and for one launch it is usually enough.

Questions and answers

Why use a spreadsheet for a launch dashboard?
It is free or cheap, familiar, flexible and quick to change. For one product it is often all you need.
How many metrics should it show?
Five to eight tiles. More makes it hard to see what matters.
How do I keep the data clean?
Use one raw data tab, one row per day or week, consistent column names and dates, and calculations on a separate tab.
How often should I update it?
Daily during launch week, then weekly. Pick a fixed time.
When should I move beyond a spreadsheet?
When you have many products, many contributors or data too large or too frequent to maintain by hand.

Sources

Ask us