Contents
- A dashboard answers a decision question, a report accounts for a period. The two share neither content nor frequency.
- Eight indicators cover almost every need: spend, cost per result, result volume, conversion rate, value generated, traffic, impression share, trend.
- A dashboard that needs a spoken explanation has missed its mark: it must read on its own, in under a minute.
- Each figure needs a reference, a target or a prior period. Without it, green or red means nothing.
- A dashboard refreshed by hand is an out-of-date dashboard: it turns back into a report.
A dashboard is not a report
The confusion costs time. A report accounts: it covers a period, explains what happened, is read once then archived. A dashboard is for deciding: it's consulted continuously, shows only what you can act on, and flags gaps rather than describing everything.
| Dashboard | Report | |
|---|---|---|
| Used for | Deciding | Accounting |
| Reading frequency | Continuous | Once, then archived |
| Content | What you can act on | What happened |
| Depth | A figure against a target | History, breakdown, commentary |
| Updates | Automatic | Per period |
The same indicator can belong to both, but not at the same depth. Cost per acquisition appears on a dashboard as a single figure compared to a target. In a report, it comes with its history, its breakdown by campaign and a commentary.
If you hesitate about a block, ask the question: can someone decide something by looking at it? If the answer is no, it belongs in the report, not the dashboard.
The eight indicators that almost always suffice
The temptation is to add. A useful dashboard works the other way: it removes until only the essentials remain.
| Indicator | What it says | Its reference |
|---|---|---|
| Spend | What's committed over the period | The planned budget |
| Cost per result | CPA or CPL depending on what you sell | The commercial target |
| Result volume | Conversions, leads, purchases | The month's goal |
| Conversion rate | Separates an acquisition problem from a site problem | The prior period |
| Value generated | Attributed revenue, or ROAS | The break-even point |
| Traffic | Sessions or clicks, to size the entry volume | The prior period |
| Impression share | Whether there's market left to take on Google Ads | 100%, and the two losses |
| Trend | The change in each of the seven above | The prior month or year |
The trend isn't one more indicator, it's a second reading of the other seven. A CPA of 45 euros means nothing; a CPA of 45 euros up 30% in a month triggers an action.
Choosing your indicators by goal
An awareness dashboard and an acquisition dashboard have almost nothing in common. Starting from the goal avoids stacking metrics by reflex.
| Goal | Primary indicator | Control indicators |
|---|---|---|
| Awareness | Reach | CPM, frequency, impressions |
| Qualified traffic | CPC | CTR, bounce rate, pages per session |
| Lead generation | CPL | Lead volume, sales conversion rate |
| Online sales | ROAS | CPA, average order value, conversion rate |
| Retention | Returning users | Purchase frequency, customer lifetime value |
An account often pursues several goals at once. In that case, several short dashboards beat one that tries everything: a dashboard has to read instantly, and instant reading doesn't survive twenty-odd indicators.
Three reading levels
A well-built dashboard reads at three speeds depending on the time available.
- Five seconds
Four figures at the top of the page, each with its change. The only question at this level: is it fine or not?
- Thirty seconds
Two trend charts. The questions become: since when, and how fast?
- Two minutes
A table by campaign or by channel. The question becomes: where exactly, and what do I change?
Most dashboards miss the first level. They open with a chart, which forces the reader to interpret before knowing whether to worry.
What makes a dashboard unusable
| Flaw | What it produces | The fix |
|---|---|---|
| Too many indicators | No one prioritizes anymore, the dashboard becomes an inventory | Cut down to twelve at most |
| Figures with no reference | A number alone doesn't inform | A target or comparison period on every figure |
| Mixed sources | A total that exists nowhere | Each figure with its source and attribution window |
| Manual updating | A dashboard current only before meetings | Connected sources, not copied exports |
On the third point: adding up Google Ads and Meta conversions produces a total that exists nowhere, the two platforms attributing over different windows. It's the subject of our article on what can't be compared between Meta Ads and Google Ads.
Building your own
- Write the question
In one sentence. "Is acquisition still profitable this month?" is a good question. "Track performance" isn't, and invariably produces a thirty-indicator dashboard.
- Derive four figures off the top
No more. These are the ones that answer the question in five seconds. Everything else is second level.
- Set a target for each figure
Even a rough one. Without a target, green or red means nothing and the reader rebuilds their own threshold, different from yours.
- Connect the sources
Rather than copying exports. A dashboard fed by hand ends up updated only before meetings, which turns it back into a report.
That's the principle Adwise takes: the platforms are connected via OAuth to the Workspace, the figures refresh on their own, and the dashboard stays accurate without intervention. The tools are detailed on the site.
Frequently asked questions
The Adwise team connects Google Ads, Meta Ads, LinkedIn Ads, GA4 and Search Console so agencies and freelancers can run their campaigns from a single place.