The essentials
- The mission: turning raw data (GA4, Ads, the CRM) into answers to business questions, not building predictive models.
- How the time splits: 40% cleaning and making data reliable, 30% dashboards and reporting, 20% analysis and recommendations, 10% training the teams.
- 2026 salaries: €38,000 to €48,000 gross early in a career, €50,000 to €65,000 with 5 years' experience; €450 to €800 a day as a freelancer.
- For a small business: a marketing platform plus a few days of consulting a year covers the need up to €200,000 to €300,000 of annual marketing budget.
The title "data analyst" has been applied to so many different roles that it no longer says much. In marketing, it means a specific function: making data usable for decisions. Neither an algorithm magician nor a mere chart producer. This article describes the job as it is practised in 2026 in marketing teams, what it costs, what separates it from neighbouring roles and how a small business can cover the need without necessarily hiring.
What a marketing data analyst actually does
The heart of the job fits in one sentence: answering business questions with reliable data. "Which channel brings the most profitable customers?", "why did the conversion rate fall in March?", "how much more can we spend on advertising before the cost per customer exceeds the margin?". To answer, the analyst goes through four activities, whose split often surprises:
| Business | Share of time | Example |
|---|---|---|
| Making the data reliable | 40 % | The measurement plan, checking the tagging, campaign naming, reconciliation with the CRM, handling the losses tied to consent |
| Building and maintaining the dashboards | 30 % | Defining the metrics with the teams, connecting the sources, layout, maintenance at every API change |
| Analysing and recommending | 20 % | Explaining the gaps, testing, one-off analyses (attribution, segmentation, profitability by product) |
| Training and sharing | 10 % | Teaching the teams to read the data, documenting the definitions |
The share spent making data reliable is the most misunderstood: a company that hires an analyst to "do analysis" discovers they spend their first six months repairing the measurement. That is normal, and it is the precondition for everything else; the method is described in auditing a GA4 account.
The everyday tools
- Sources: GA4, Search Console, Google Ads, Meta Ads, LinkedIn Ads, the email platform, the CRM, sometimes the till software or the ERP.
- Collection and storage: Google Tag Manager, connectors, BigQuery or an equivalent warehouse for large volumes; see BigQuery and GA4.
- Processing: SQL first, advanced spreadsheets, Python for one-off analyses.
- Presentation: Looker Studio, Power BI, or an integrated marketing platform.
- AI: assistants for writing SQL, summarising a report, generating a first commentary; the analyst checks and interprets.
Data analyst, data scientist, marketing ops: who does what
| Role | Typical question | Dominant skill | Relevant to a small business? |
|---|---|---|---|
| Marketing data analyst | "What happened and why?" | SQL, dashboards, knowledge of the marketing platforms | Yes, often part-time or outsourced |
| Data scientist | "What is going to happen? Which model predicts best?" | Statistics, machine learning, Python | Rarely before several million rows of data |
| Marketing ops | "How do the tools talk to each other and who triggers what?" | Automation, CRM, integrations | Yes as soon as a CRM and sequences exist |
| Traffic manager | "How do we optimise this campaign?" | The advertising platforms | Yes, in-house or at an agency |
Salaries and rates in 2026
In France, a junior marketing data analyst earns €38,000 to €48,000 gross a year, a profile with 3 to 5 years' experience €50,000 to €65,000, more in Paris and in scale-ups. The full employer cost of a mid-level profile is around €75,000 to €90,000 a year including tools. Freelance day rates run from €450 to €800, with setup engagements typically lasting 5 to 15 days then 1 to 2 days a month of follow-up. The support formats and their costs are compared in what analytics support costs.
Hire, outsource or buy a tool: the threshold for a small business
- A marketing budget under €100,000 a year: a marketing platform at €39 to €250 a month that automates the collection and the dashboards, plus 2 to 5 days of consulting a year for the measurement plan and a quarterly reading. A full-time analyst would cost more than the budget they optimise.
- A budget of €100,000 to €300,000: the same tooling, plus a freelancer 1 to 2 days a month for the analysis and the recommendations, or a marketing person trained to read the data; see training your teams to analyse KPIs.
- Above €300,000, or several brands and countries: an in-house analyst is justified, provided you give them a collection tool already in place so they don't spend their time re-entering data.
In all three cases, the classic mistake is hiring before the measurement is reliable and the collection is tooled: the analyst then becomes a producer of manual reports, and the expected value never arrives.
How GreenRed helps
Rather than juggling several tools, GreenRed's Continuing training module brings these metrics together in a single dashboard, compares them over time and tells you which actions come first. You can try it free, with no card, from the Pricing.
Frequently asked questions
Does a marketing data analyst need to be able to code?
SQL is essential for querying a data warehouse or BigQuery. Python is useful for one-off analyses but not mandatory in a small business. The rarest skill isn't the code: it is the ability to connect a figure to a business decision and explain it in plain language.
What training does a marketing data analyst need?
The profiles come from business schools with a data specialisation, university courses in statistics or quantitative marketing, or career changes through intensive 3- to 6-month courses. The GA4 and Google Ads certifications, which are free, remain a practical prerequisite for marketing.
Can an apprentice hold this role in a small business?
An apprentice can build and maintain dashboards and produce simple analyses, if supervised by someone who knows the data and if the collection is already automated. Giving them the measurement plan with no supervision leads to wrong data for months.
Will AI replace the data analyst?
AI speeds up writing queries, formatting and a report's first commentary; it cuts production time by 30 to 50%. It doesn't verify that the data is right and doesn't know the company's context. The role shifts towards making data reliable, interpreting it and deciding.