TL;DR
- A dashboard is a display layer that pulls live data from systems such as GA4 and Google Ads; it does not store the data itself.
- Dashboards are commonly classed as strategic (executives, few KPIs), operational (daily teams, frequent refresh) and analytical (filters for finding causes).
- Google Ads and GA4 count conversions under different rules, so a small gap is normal and a large one needs tracing before display.
- Put the biggest result number top left, add a target or prior period for context, and write a short definition beside any custom count.
A dashboard is a single screen that gathers the key numbers in one place so the viewer can see the state of a business or campaign at a glance. It usually shows large figures, charts and short tables fed by live data. It differs from a report in that a dashboard is built to be opened repeatedly to answer "how are we doing right now", while a report is usually a document that summarises a period and explains it.
What a dashboard is, and how it differs from a report
A dashboard is a display layer for data, not a place where data is stored. The numbers are pulled from other systems such as GA4, Google Ads, Search Console, a CRM or a spreadsheet, then arranged so they are easy to read. If the source data is wrong, a beautiful dashboard simply shows wrong numbers beautifully.
The difference in function
A dashboard answers status questions, such as "where is this month's cost per lead against target", and it is opened often. A report answers explanatory questions, such as "why did cost rise last month and what do we do next", which needs a person to interpret the data and write conclusions. In practice the two work together: the dashboard shows that a number moved, and the report explains why.
| Aspect | Dashboard | Report |
|---|---|---|
| Main job | Show current status, repeatedly | Summarise and explain a period |
| How often opened | Often, possibly daily | On a cycle, such as monthly or quarterly |
| Interpretation | Little; the viewer interprets | Yes; the author adds conclusions and recommendations |
| Data freshness | Follows the system's refresh cycle | Fixed as of the date it was written |
Types of dashboard
A widely used way to classify dashboards is by user and purpose, usually into three kinds: strategic, operational and analytical. This is a thinking framework, not a mandatory standard, and one real dashboard can blend them.
Strategic dashboards
These serve executives who want a high-level view. They carry few numbers, mostly headline KPIs such as revenue, cost per customer acquired and progress against target. They do not need frequent updates, perhaps weekly or monthly, because the aim is to see direction, not to chase daily problems.
Operational dashboards
These serve teams running day-to-day work, such as the person managing ads. Numbers refresh often, for example spend so far today, the number of conversions and campaigns close to their budget limit. The purpose is to catch anomalies early enough to fix them.
Analytical dashboards
These serve people who need to trace causes. They offer filters and dimensions to drill into, for example by campaign, device or period. They usually hold more detail than the other two and expect more skill from the reader.
| Type | Main user | Nature of the numbers |
|---|---|---|
| Strategic | Executives | A few headline KPIs, showing direction |
| Operational | Teams running daily work | Frequent updates, focused on catching anomalies |
| Analytical | Analysts and account managers | Filters and dimensions to find causes |
What metrics a marketing dashboard should include
There is no list of metrics that is right for every business. A workable principle is to start from the questions the viewer needs answered and pick only the numbers that answer them. The layout below is a common starting idea, not a required set.
Top row: result numbers
Put the numbers tied to the business goal at the top and make them the largest, such as leads or conversions, cost per lead and ROAS if there are online sales, each compared with target or the previous period. The viewer should be able to tell good from bad without doing arithmetic.
Middle row: numbers that explain the result
These help explain why results moved, such as clicks, CTR, CPC and landing page conversion rate, broken out by channel, for example search ads, organic search and social.
Bottom row: detail to drill into
Tables by campaign, page or search term for people who want to go deeper. Place this at the end of the page or on a separate page, so executives are not faced with numbers that do not bear on their decisions.
What to cut
Numbers that have never changed a decision should leave the main page. Totals such as overall impressions or follower counts can sit on a secondary page but should not take prime space if the business does not decide from them. Numbers that do not match the agreed KPI definition should also come off.
Dashboard tools, without brand cheerleading
Several kinds of tool are in common use. Choose by the data sources you must connect, your team's skills and who will open the result, not by which tool is better known. What follows is general principle, and you should always check each tool's current features and terms before deciding.
Looker Studio
Looker Studio is Google's dashboard tool, formerly named Google Data Studio. It has connectors for GA4, Google Ads, Search Console and Google Sheets, so it suits marketing teams whose data mostly lives in Google's systems. The caution is that data from several systems without a ready-made connector usually has to be prepared first.
GA4 reports and Explorations
GA4 has built-in reports and an Explorations tool, which suit looking at web and app data without setting up another tool. They are best for analysing GA4's own data. If you want to lay ad data or customer data from other systems alongside it, you usually need an additional tool.
Power BI
Power BI is a Microsoft tool that supports a wide range of data sources and more complex data modelling. Businesses already on Microsoft systems, or with a data team, often choose it.
| Tool | Mechanical strength | Check before choosing |
|---|---|---|
| Looker Studio | Connectors to Google services; easy to build and share | Sources outside Google need data preparation |
| GA4 Reports and Explorations | Sits in the same system as the web data; no extra setup | Limited ability to combine other systems |
| Power BI | Supports many sources and complex models | Check data source support, licensing and current terms |
Design principles for a dashboard that is easy to read
A good dashboard lets the viewer answer the main question within a few seconds. The common problem is not the tool but too many numbers, so the viewer does not know where to look.
Set priority with position and size
A common design convention is to put the most important number at the top left and make it the largest. Secondary numbers go below and smaller. Using size and position rather than loud colour makes the priority clear without clutter.
Match the chart to the question
Line charts suit trends over time, bar charts suit comparing categories, and a single large number suits a value that needs an instant read, such as this month's cost per lead. Avoid charts that are hard to interpret without need, because viewers then spend time decoding instead of deciding.
Give numbers context
A lone number says little. Add a comparison such as a target, the previous period or the same period last year, and label the period shown clearly. If a number uses a specific count, such as only qualified leads, write a short definition near it.
Cut unneeded screens and filters
If the main page needs long scrolling or dozens of filters, people stop opening it. Have someone who did not build the dashboard open it and say what they see. If they cannot state the intended message within seconds, adjust it.
Combining several data sources into one dashboard
Combining data is the hardest part of a marketing dashboard, because the data sits in several systems and each defines numbers differently. For example, Google Ads and GA4 count conversions under different rules. A small difference is normal, but a large one needs tracing before the number is displayed.
An orderly sequence
- Agree the KPI list and the definition of each, including which system is the primary source.
- Check that collection at the source is correct, for example that GA4 key events match what the business treats as valuable.
- Choose how to combine the data: ready-made connectors, a central table or a data warehouse, depending on size and complexity.
- Join the data on matching keys, such as lead source or a campaign identifier carried in UTM links.
- Test the dashboard against source-system numbers before publishing, and recheck periodically.
Common problems
The main problems are ad links without consistent UTM parameters, which mixes up sources; lead status in the sales system not being sent back, which makes quality unmeasurable; and time zones or currencies that differ between systems. Fixing these at the source is cheaper than patching them in the dashboard.
If your data is scattered across systems and you want the team to work from one set of numbers, the data integration and consolidation service is one starting point, and if GA4 measurement needs to be set up correctly before a dashboard, see the Google Analytics 4 migration and setup service. For organisations planning a wider data foundation, see the digital transformation service.
Example: a search ads dashboard layout for a service business (illustrative)
This is a hypothetical layout to show the thinking. It does not come from a real account and carries no benchmark figures.
- Top row: qualified leads this month against target and last month, and cost per qualified lead.
- Middle row: a line chart of daily leads and ad spend, and a table separating brand campaigns from non-brand campaigns.
- Bottom row: a table of search terms or ad groups that spend heavily but produce no leads, for the account manager to decide budget moves.
- Note under the numbers: the definition of "qualified lead", the data source and when the data last refreshed.
The layout answers in order: how are results, what produced them, and what should be fixed. An executive can read only the top row, while an account manager uses all three. For teams that want the ad account manager to set up measurement and reporting aligned with the goal, see the Google Ads service.
What this means for Thai marketers
This section is analysis, not a claim that a source studied the Thai market. Many Thai businesses win customers through several channels at once, such as forms, phone calls and chat in LINE, so data is scattered and a dashboard drawing on one system shows an incomplete picture.
What to check in your own account
- Ask what question your current dashboard answers and who really opens it. Remove numbers nobody looks at.
- Check whether the channels customers actually use, such as calls and LINE, are recorded as measurable events yet.
- Compare conversion numbers between Google Ads and GA4 and note the cause of any gap.
- Write the definition of each number on the dashboard itself so it cannot be read two ways.
- Check when the data last refreshed before using a number to decide.
If you want specialists to help set up data collection, combine sources and build a dashboard your team will really use, you can talk to the Relevant Audience team through the related services above.
Frequently asked questions about dashboards
What is a dashboard in the shortest possible terms?
A dashboard is one screen that gathers key numbers pulled from live data so you can check the status of work quickly. It is not itself a store of data.
How is a dashboard different from a report?
A dashboard is built to be reopened to see current status, while a report summarises a period with the author's explanation and recommendations.
How many numbers should a marketing dashboard have?
There is no fixed count, but in principle it should hold only what answers the viewer's main questions, with the most important result numbers first and explanatory and detail numbers following in order.
Should I use Looker Studio, GA4 or Power BI?
It depends on your data sources and your team's skills. If most data is in Google's systems, Looker Studio usually connects easily, and if you need complex data models Power BI is a common choice. Check current features and terms before choosing.
Why don't dashboard numbers match the source system?
Common causes are different counting definitions, mismatched periods or time zones, data still processing, or different filters. Compare them one by one before concluding which number is wrong.
