A single marketing and sales dashboard shows, on one page, what acquisition costs, how many enquiries it produces and how many of those enquiries turn into revenue. Building one takes three ingredients: reliable sources joined by a shared key (campaign source, lead, order), a short list of KPIs defined in writing, and a visualisation tool that fits your volume, such as Looker Studio, Power BI or your CRM's reporting module.
The main pitfall is not technical. It is piling up charts without deciding which question each one answers. A good dashboard fits on one screen, can be read in five minutes during the weekly meeting and leads to decisions. This guide covers the sources to connect, the KPIs worth keeping, tool choices and the most common mistakes.
Why one dashboard
In a small business, marketing usually looks at Meta Ads and GA4, sales looks at the CRM and management looks at the accounts. Everyone has their own figures, and meetings are spent arguing about which ones are right. A shared dashboard lets you:
- connect ad spend to revenue actually collected, not just to clicks;
- see where the journey stalls: too few enquiries, poorly qualified enquiries, or enquiries handled badly;
- share a vocabulary, so a "lead" means the same thing to everyone.
Prerequisites
Before connecting anything, make sure you have:
- clean web measurement (GA4 key events, consistent UTMs);
- a CRM the team actually uses, where every lead has a source and a stage;
- a single source of truth for sales: ERP, invoicing software or e-commerce back office;
- a KPI dictionary: name, definition, formula, source, owner;
- one person responsible for the dashboard, who fixes data breaks.
Without these, the dashboard will display wrong numbers very elegantly.
Sources to connect
| Source | What it provides | Usual connection | Watch out for |
|---|---|---|---|
| GA4 | Sessions, sources, key events | Native Looker Studio connector, API | Gaps caused by consent and ad blockers |
| Search Console | Queries, impressions, organic clicks | Native Looker Studio connector | Aggregated data; some queries are anonymised |
| Google Ads | Cost, clicks, conversions | Native Looker Studio connector | Google Ads' own attribution |
| Meta Ads | Cost, reach, leads | Third-party connector or export | Attribution windows differ from GA4 |
| CRM (HubSpot, Brevo, Zoho, Odoo…) | Leads, stages, won deals | Connector, API, scheduled export | Data entry quality |
| Invoicing / e-commerce | Orders, revenue, refunds | Export, database, API (WooCommerce, Shopify) | Returns and cancellations |
For sources without a native connector, you have two options: a scheduled export to Google Sheets or a database, or an automated flow in n8n or Make that feeds an intermediate table. The second needs a little more maintenance but removes manual copy-paste.
Choosing the right KPIs
Keep it to eight to ten indicators, organised along the customer journey.
Acquisition
- Media spend by channel.
- Qualified sessions, or visitors reaching a key page (offers, pricing, contact).
Conversion
- Leads (quote requests, calls, sign-ups) by source.
- Cost per lead = spend / leads, by channel.
- Qualification rate = qualified leads / total leads.
Sales
- Deals won and lead-to-customer conversion rate.
- Revenue attributed to sources, with the attribution rule stated.
- Average first-response time and sales cycle length.
Profitability
- Customer acquisition cost (CAC) = marketing and sales spend / new customers, over the same period.
- For e-commerce: average order value and ROAS (revenue / ad spend), keeping in mind that ROAS ignores margin.
Give each KPI a target, or at least an expected trend. A number with no reference point triggers no decision.
Choosing the tool
| Tool | Strengths | Limits | Best for |
|---|---|---|---|
| Looker Studio | Free, native Google connectors, easy sharing | Slow with many sources; third-party connectors often paid | Businesses centred on Google tools |
| Power BI | Strong data modelling, Microsoft 365 integration | Learning curve; licences needed for sharing | Companies already on Microsoft |
| CRM reporting | No integration needed for sales data | Little external marketing data | Sales-first tracking |
| Spreadsheet (Google Sheets, Excel) | Flexible, familiar | Fragile, formula errors, weak access control | Pilot phase, low volumes |
Starting in a spreadsheet to validate the KPIs, then moving to Looker Studio or Power BI once definitions are stable, is a sensible path.
Building it in 5 steps
- Write down the questions it must answer ("Which channel brings us customers at the most reasonable cost?").
- Define the KPIs in the dictionary, with formulas.
- Connect and clean the sources: consistent campaign names, currencies, time zones.
- Sketch one page: headline numbers at the top, trends in the middle, channel detail at the bottom.
- Test with the people who use it for two or three meetings, then adjust.
Example for a Tunisian SME
Consider a B2B supplier of professional equipment that sells through quote requests and runs Meta and Google Ads campaigns. Its page could show, at the top, the month's spend, number of requests, cost per request and signed revenue; in the middle, weekly requests by channel; at the bottom, a table per campaign linking spend, requests, quotes sent and deals won. The link between campaign and sale relies on the source recorded in the CRM when the request comes in.
Pitfalls to avoid
Mixing definitions
Meta Ads "conversions", GA4 key events and CRM leads are not counted the same way. Show them separately and name the source under each number.
Combining incompatible periods or attribution windows
A sale attributed by Google Ads over a 30-day window and a CRM sale recorded at signature do not necessarily land in the same month. State clearly which date each figure uses.
Ignoring data entry quality
If a large share of CRM leads have no source, the dashboard cannot connect marketing to sales. Make the source field mandatory and check it weekly.
Too many pages
Beyond three or four pages, nobody opens them. Keep one summary page and send people to the source tools for detail.
Exposing personal data
A shared dashboard rarely needs names or phone numbers. Aggregate, restrict access and comply with the data protection rules that apply to you, including Tunisia's Organic Law No. 2004-63.
Forgetting maintenance
An expired connector, a renamed campaign or a new form is enough to break a chart. Schedule a monthly check and keep a change log.
Running the weekly review
A dashboard only earns its keep if it is used in a fixed ritual. A simple format works well: the same day each week, fifteen minutes, the same page on screen. Start with the headline numbers against their targets, then look at the one or two indicators that moved most, and finish by agreeing on concrete actions with an owner and a date: pause a campaign, call back unanswered enquiries, fix a form. Record those decisions next to the dashboard so the following week starts by checking whether they were carried out. Once a quarter, review the KPI dictionary itself: drop indicators nobody used, and add one only if it answers a question the team keeps asking.
Limits to keep in mind
A dashboard does not replace analysis or field knowledge. Attribution remains an estimate: part of your sales will always come from word of mouth, trade fairs or direct calls, which digital tools capture poorly. And gaps between platforms are normal; the goal is to know and explain them, not to make them disappear.