A dashboard that tries to answer everything answers nothing. The ecommerce dashboards that work each have one question, a few metrics and a layout that leads the eye from the number to the cause. This guide offers five templates (sales, products, customers, marketing, leadership), each with its question, metrics, recommended layout and the typical decision that comes out of it, then explains how to build the hierarchy and how to review a dashboard you already have.
Why five dashboards with one question each
One question, one view. If a screen mixes sales, campaigns, customers and inventory, whoever opens it does not know where to start and ends up looking at the biggest number. With five views, each with its own question, you start from the first and drill down where needed.
An example with made-up numbers. Last month a store made €60,000 from 1,000 orders. This month it made €56,400 from 940 orders: down 6%. The sales dashboard shows it at once: average order value is identical (60,000 ÷ 1,000 = €60; 56,400 ÷ 940 = €60), so the problem is orders, not how much each buyer spends. The next question sits in the marketing dashboard: sessions are flat at 40,000, but conversion went from 2.5% to 2.35% (940 ÷ 40,000). Traffic did not fall. The share that buys did. The next step is working out which source and which device, and then looking at products. Three views, three questions, one conclusion. On a single screen with thirty charts the same path would be a treasure hunt.
The five templates
Each template has a name, a question and a layout. The layout is always the same: big numbers at the top, a chart in the middle, a table at the bottom to drill into the detail.
1. Sales dashboard
Question: how are sales doing compared with before, and what do they depend on?
Metrics (5): net sales, orders, average order value, conversion rate, sales by channel or source.
Layout: three big numbers at the top (sales, orders, average order value), each with the comparison against the previous period. A line chart in the middle with the current week and the previous one. A table by channel at the bottom.
Typical decision: if average order value is steady and orders fall, the problem is traffic or conversion and you move to the marketing view; if orders are steady and average order value falls, look at discounts and product mix.
2. Products dashboard
Question: which products carry the store, and which need rethinking?
Metrics (5): sales by product, share of the top three products in the total, units sold, refunds by product, stock availability.
Layout: one big number with the share of the top three products. A sorted bar chart of sales. A table with product, sales, refunds and availability.
Typical decision: a product that sells a lot but has many refunds gets fixed before it is restocked; one that has sold little for months is a candidate to drop or relaunch; a high top-three share tells you how much the store depends on a few items.
3. Customers dashboard
Question: who buys, and who comes back?
Metrics (5): new customers, returning customers, returning share, time between first and second order, average order value for new and returning customers.
Layout: the returning share at the top. A chart in the middle with new and returning customers over time. A table at the bottom of the products people start with and how many of those customers buy again.
Typical decision: if the average time between first and second order is 40 days, a repeat-purchase message makes sense around that point, not after three months; entry products that drive the most repeat orders deserve more space.
4. Marketing dashboard
Question: which sources bring orders that are worth having, and at what cost?
Metrics (5): sessions by source, conversion by source, orders by source, spend by source, customer acquisition cost.
Layout: acquisition cost and conversion at the top. A bar chart by source in the middle. A table at the bottom with source, sessions, orders and spend.
Typical decision: a source with many visits and few sales is not cut automatically: you check whether the audience is the right one. A source with rising cost and low repeat purchase is the candidate for a smaller budget. Spend is not in the store's own data: it has to come from other sources.
5. Leadership dashboard
Question: are we growing in a healthy way?
Metrics (5): net sales over the last twelve months, share of returning customers, share of top products, refund rate, margin if available.
Layout: a few big numbers, a twelve-month trend chart, a small table of exceptions. No operational detail.
Typical decision: if sales grow but the returning share falls and dependence on top products rises, the growth is fragile, and you invest in assortment and loyalty before buying more traffic.
How to build the hierarchy
The hierarchy is the order in which the eye reads. Three levels are enough.
- Level 1: the number. At most three big numbers, each with its comparison. It is what you read in five seconds.
- Level 2: the why. A chart that shows the trend. It helps you see whether it is a spike or a direction.
- Level 3: the detail. A table sorted by the most important value, to drill down to the product, customer or source.
Two practical rules. Always sort by importance, not alphabetically. And put at the top left whatever you decide on first.
A checklist for reviewing a dashboard you already have
- Can you state the question it answers in one sentence? If it takes two, it may be two dashboards.
- Does every number have a comparison (previous period or your recent average)?
- Are there more than six metrics per view? Remove the ones nobody decides on.
- Is the most important number the most visible?
- Are the definitions written down (what counts as sales, who is a returning customer)?
- Has anyone changed a decision in the last month by looking at this view? If not, rework it.
- Does whoever opens it know what to do next?
Frequent mistakes
| Mistake | Why it hurts | How to fix it |
|---|---|---|
| Too many charts in one view | Nobody knows where to start | One question per view, six metrics at most |
| Numbers without a comparison | You can't tell if the value is good | Previous period or your average over recent weeks |
| Averages instead of segments | The average hides the product or channel getting worse | A table by product, channel and customer type |
| The wrong chart for the question | A pie with twelve slices can't be read | Lines for time, sorted bars for comparing |
| Definitions that change | The meeting discusses the numbers, not the store | Write the definitions down once |
Which template to start from
| Store stage | Start with | Add later |
|---|---|---|
| You have just opened | Sales and products | Customers, once the first repeat orders arrive |
| Growing with advertising | Sales and marketing | Products, to see where to push |
| You have a customer base | Customers and products | Marketing, to keep costs in check |
| You report to others (partners, leadership) | Leadership | The rest, only for the questions it raises |
How using a dashboard changes when you can also ask your data questions
A dashboard answers the questions someone anticipated. When a new one comes up, for example "which products fell compared with last month?", it used to mean a new chart. With a tool you can ask in plain language, you start from the summary view and dig in with a question. The work changes like this: the dashboard stays the place where you work out what to look at, and the questions take you to the why. Remember that an automatically generated answer should be checked against the view and the underlying data before you decide.
Where Opalca View fits
Opalca View is a set of ready-to-use dashboards for sales, products and customers, with automatic weekly and monthly reports: they cover the same themes as the first three templates in this guide, already set up. The connector that is live today is Shopify; others are on the way. The Growth and Enterprise plans include Ask Opalca, to ask questions of your store's Shopify data, refreshed at the frequency of your plan. Ask Opalca does not know advertising spend, which is not in Shopify data. The marketing and leadership templates often need extra sources: if you need them, a custom project can be assessed with Opalca Lab.
Have a question or want to talk it through? Contact us. You can see the plans on the Pricing page.
Keep reading
- Ecommerce KPIs: the 10 metrics that matter and how to read them
- Shopify reporting: how to automate your store reports
Frequently asked questions
How many dashboards does an ecommerce business need?
Two is a good start: sales and products. The others are added when they serve a decision.
How many metrics per dashboard?
Four to six. If a metric leads to no decision, it is a statistic and comes out.
One dashboard or several?
Several, each with one question. A single view only works with very few numbers. For everything else, a summary at the top and the detail in separate views reads better.
Which chart should I pick?
Lines for trends over time, sorted bars for comparing products or sources, a table for the detail. The big number is for what you read first.
Are an AI's answers on my data reliable?
They are useful for getting oriented, but they are generated automatically: check them against the dashboard and the underlying data before an important decision.