Skip to content
Contact us

Measurement and analytics

Website monitoring

If the checkout form stops working, the analytics code is removed by mistake or robots.txt blocks …

Read more

Tool setup

Among other things, we use analytics tools to measure how areas such as SEO and conversion …

Read more

Web analytics

Everything we do starts with analysis. A web analysis helps us understand:

Read more
On this page

Without measurement, a company has to guess how its digital marketing is performing. Measurement and analytics show what works, what does not and where the next effort should go.

Analysing visitor behaviour and traffic shows how a website can attract more visitors or how its conversion rate can be improved. The work consists of collecting data, analysing it and turning the insights into concrete actions.

The right analysis answers key questions: Where does the traffic come from? Which pages perform best? Where in the conversion funnel do users drop off? Which segments convert best? Is the SEO work delivering results? Which ad campaigns have the best ROI?

With tools for reporting and statistical analysis, a company can track behaviour, identify segments and optimise its efforts based on actual results rather than assumptions.

Why measurement and analytics are critical

Eliminate guesswork

The problem without analysis: decisions are made on gut feeling or assumptions.

Examples

  • “We think SEO is working well” (but no data confirms it)
  • “This landing page feels good” (but nobody knows the conversion rate)
  • “Facebook ads ought to deliver results” (but nobody has done an ROI analysis)

With analysis: Facts drive the decisions. If SEO is not delivering traffic, it shows in the data. If a landing page converts poorly, that shows too, and Facebook ads with a negative ROI are stopped.

Identify opportunities

Analysis reveals both problems and opportunities

Examples

  • Organic traffic from a specific search phrase is growing fast → Invest more in related search phrases
  • One product page converts five times better than the others → Find out why and apply it to other pages
  • Email campaigns to segment X have a 20 per cent open rate compared with 5 per cent for segment Y → Focus on segment X

Measure the ROI of your efforts

Every marketing effort costs money

  • SEO (time, content production)
  • Advertising (direct cost)
  • Conversion optimisation (development, design)

Analysis shows what delivers a return

Examples

  • An SEO project cost SEK 50,000 and generated SEK 200,000 in contribution margin → ROI 300 per cent
  • Facebook ads cost SEK 30,000 and generated SEK 15,000 in contribution margin → ROI minus 50 per cent (a loss)

Result: Invest more in SEO, and reduce or stop the Facebook ads.

Improve continuously

Analysis is recurring work: measure, analyse, act and measure again.

Example

  1. Measure the conversion rate: 2 per cent
  2. Analyse: Where do users drop off?
  3. Action: Shorten the forms
  4. Measure again: the conversion rate is now 2.5 per cent
  5. Analyse: What else can be improved?
  6. Repeat

Result: Long-term growth through continuous optimisation.

Which types of analysis are needed?

Traffic analysis

Purpose: Understand how the number of visitors from different sources is developing.

What is analysed

  • Total traffic over time (growing, shrinking or stable?)
  • Traffic by source (organic search, paid search, direct, social, referring websites)
  • Traffic by channel subcategory (which search phrases, which ads, which social platforms)
  • Geographical distribution (where do visitors come from?)
  • Device distribution (desktop, mobile, tablet)

Use

Tracking SEO work
Does organic traffic increase after SEO work? From which search phrases?
Tracking ad campaigns
Are Google Ads and Facebook Ads delivering traffic? At what cost per visitor?
Identifying trends
Seasonal variation? Sudden peaks or drops? Why?

Behavioural analysis

Purpose: Understand what visitors do on the website.

Behavioural analysis can serve different purposes

1. Laying the groundwork for developing the website’s structure

What is analysed

  • Which pages are visited most?
  • From which pages do most visitors leave the website (exit rate)?
  • Which pages have the highest bounce rate (the share of sessions that were not engaged)?
  • How do users navigate between pages?

Use: Restructure the navigation, improve internal links, identify important pages that should be optimised.

2. Segmenting visitors

What is analysed

  • Which visitors are likely to become customers?
  • Which visitors are not ready to convert in the short term?
  • Which behaviours indicate strong purchase intent?

Use: Create segments for targeted communication (remarketing, email).

3. Pinpointing weak links in the conversion chain

What is analysed

  • Conversion funnel: Where do most users drop off?
  • Which pages before the conversion have problems?
  • Which form fields cause abandonment?

Use: Conversion optimisation, A/B testing, action plans.

4. Segmenting visitors for ad sales

What is analysed

  • Behaviour patterns among different visitor segments
  • Demographics and interests (if the data exists and the visitor has given consent)

Use: Media companies or platforms can sell ad space with better audience targeting based on behavioural data. The data collection requires a legal basis and, for cookies and similar technologies, consent.

5. Analysing how visitor segments respond to ads

What is analysed

  • Which visitor segments click on ads?
  • Which ad content works for which segments?
  • Which ad placement gives the best results?

Use: Optimise display ads, native ads and sponsored content for maximum engagement and revenue.

Conversion analysis

Purpose: Understand what drives (or prevents) conversions.

What is analysed

  • Conversion rate overall and by traffic source
  • Conversion funnel (which steps, where people drop off)
  • Micro-conversions (steps on the way to a macro-conversion)
  • Attribution (which channels are credited with the conversion?)

Use: CRO (Conversion Rate Optimisation), improving landing pages, optimising the checkout.

SEO analysis

Purpose: Measure the effect of SEO work.

What is analysed

  • Rankings for target search phrases
  • Organic traffic over time
  • Clicks, impressions and CTR per search phrase (Google Search Console)
  • Indexed pages and indexing problems (the Page indexing report)
  • Backlinks (number, quality)

Use: Understand whether the SEO work is delivering results, identify new opportunities, find technical problems.

Ad analysis

Purpose: Measure the ROI of paid traffic.

What is analysed

  • CPC (cost per click)
  • CTR (click-through rate)
  • Conversion rate from ads
  • CPA (cost per acquisition)
  • ROAS (return on ad spend)

Use: Optimise ad campaigns, pause underperforming ads, scale up winning campaigns.

Competitor analysis

Purpose: Understand what competitors are doing and how your own performance compares.

What is analysed

  • Competitors’ rankings for the same search phrases
  • Competitors’ estimated traffic
  • Competitors’ backlinks
  • Competitors’ advertising strategies

Use: Find gaps (search phrases where competitors rank but the company does not) and new opportunities.

How the analysis process works

Step 1: Define goals and KPIs

The goals must be clear before the analysis begins.

Examples of goals

  • Increase organic traffic by 30 per cent in six months
  • Improve the conversion rate from 2 per cent to 3 per cent
  • Reduce CPA from SEK 500 to SEK 350
  • Increase email subscriptions by 50 per cent

KPIs (Key Performance Indicators): Measurable values that show whether goals are being met.

Examples

  • Organic traffic per month
  • Conversion rate
  • Cost per acquisition (CPA)
  • Email subscriptions per month

Step 2: Set up tracking

Data must be collected correctly.

Consent: Cookies and similar technologies require the visitor’s consent unless they are needed to deliver the service. Processing personal data also requires a legal basis under the GDPR. Consent Mode adapts Google’s tags to the visitor’s choice. The analytics tool Matomo can be run on your own server, which gives you control over where the data is stored.

The basics

  • Google Analytics 4 installed
  • Key events defined
  • Event tracking for important actions (clicks, forms, purchases)

Advanced

  • Google Tag Manager for flexible tracking
  • E-commerce tracking
  • Custom events
  • Cross-domain tracking (if there are several domains)

Without correct tracking, all analysis is worthless.

Step 3: Collect data

Let data accumulate over a sufficiently long period.

Minimum period

  • Traffic analysis: one month (preferably at least three months for trends)
  • Conversion analysis: until there are enough conversions to assess a difference (the number depends on the size of the effect and the chosen significance level)
  • SEO analysis: three to six months (SEO takes time)

Step 4: Analyse the data

Ask the right questions

What happened? Descriptive analysis. Traffic increased by 20 per cent.

Why did it happen? Diagnostic analysis. Traffic increased because organic traffic from a specific search phrase grew.

What will happen? Predictive analysis. If the trend continues, the goal will be reached in three months.

What should be done? Prescriptive analysis. Invest more in content for that search phrase and optimise more related pages.

Step 5: Report the insights

The analysis must be communicated to decision-makers.

A good report has

  • A clear summary (executive summary)
  • Visualisations (graphs, tables)
  • Concrete insights (not just figures)
  • Recommendations (what should be done?)

Avoid

  • Data dumps (200 slides of figures without context)
  • Jargon (explain the terms)
  • Insights without action (what should be done with the information?)

Step 6: Act on the insights

An analysis is only useful once it leads to action.

From insight to action

Insight
Organic traffic from search phrase X is growing fast.
Action: Create more content around related search phrases.
Insight
The checkout has a 70 per cent drop-off at the payment step.
Action: A/B test a simplified payment form.
Insight
Facebook ads have a CPA of SEK 800 compared with SEK 400 for Google Ads.
Action: Reduce the Facebook budget and increase the Google Ads budget.

Step 7: Measure again and iterate

A continuous cycle: Implement actions → Measure results → Analyse → Next action.

Example

  1. Measured conversion rate: 2 per cent
  2. Action: Shortened the forms
  3. Measured again: conversion rate 2.3 per cent
  4. Action: Added trust signals
  5. Measured again: conversion rate 2.7 per cent
  6. Carry on in the same way

Tools for measurement and analytics

Analytics platforms

Google Analytics 4 (GA4)

  • Visitors, sessions, page views
  • Traffic sources
  • Path exploration (the routes users take)
  • Key events
  • E-commerce tracking
  • Real-time data

Adobe Analytics

  • Enterprise-level analytics
  • Advanced segmentation
  • Predictive analytics

Tracking and tag management

Google Tag Manager

  • Implement tracking without code changes
  • Event tracking
  • Custom variables
  • Integrate third-party tools

SEO tools

Google Search Console

  • Search phrases (impressions, clicks, CTR, position)
  • Indexed pages
  • Indexing problems (the Page indexing report)
  • Core Web Vitals
  • Backlinks

Semrush

  • Competitor analysis
  • Keyword research
  • Backlink analysis
  • Site audit

Ahrefs

  • Backlink analysis
  • Keyword research
  • Content explorer
  • Rank tracking

Advanced Web Ranking (AWR)

  • Rank tracking for search phrases
  • Follow rankings over time
  • Competitor comparison

Rank Tracker

  • Monitor keyword positions
  • Local SEO tracking

Website Auditor

Advertising platforms

Google Ads

  • Search advertising
  • Display
  • Shopping
  • Performance tracking (impressions, clicks, conversions)

Facebook Ads Manager

  • Ads on Facebook and Instagram
  • Audience targeting
  • Performance metrics

Meta Business Suite

  • Analytics across several platforms (Facebook, Instagram)
  • Audience insights

Visualisation and reporting

Data Studio

  • Build your own dashboards
  • Combine data from several sources (GA, Search Console, Ads)
  • Automated reports
  • Free

Tableau

  • Advanced data visualisation
  • Enterprise BI

Power BI

  • Microsoft’s BI tool
  • Integration with Microsoft 365

Heatmaps and session recordings

Contentsquare (formerly Hotjar)

  • Heatmaps (clicks, scrolling)
  • Session recordings
  • Surveys

Microsoft Clarity

  • Free heatmaps and recordings (requires a consent signal for visits from the EEA, the UK and Switzerland)
  • Rage clicks, dead clicks

A/B testing

Tests in existing systems

  • Modern websites are complex, so in our experience external testing tools are rarely reliable on their own
  • The test is usually built into the company’s existing systems

Separate analysis

  • Data from the test is extracted for analysis outside the test environment
  • Each visitor’s variant is linked to the outcome

Customer data platforms

Segment

  • Centralise data from different sources
  • Send data to analytics, CRM and marketing tools

Which reports are needed?

Traffic report

Content

  • Total traffic over time (graph)
  • Traffic by source (table or pie chart)
  • The top landing pages
  • Geography
Frequency
Monthly or weekly.
Audience
Marketing team, management.

SEO report

Content

  • Rankings for target search phrases
  • The trend in organic traffic
  • Newly indexed pages
  • Technical problems (if any)
  • Backlinks acquired
Frequency
Monthly.
Audience
SEO team, marketing team.

Advertising report

Content

  • Ad spend per campaign
  • Impressions, clicks, CTR
  • Conversions, CPA
  • ROAS
  • Recommendations (scale up or pause campaigns)
Frequency
Weekly or monthly.
Audience
Marketing team, management.

Conversion report

Content

  • Conversion rate overall and by source
  • Funnel analysis (where people drop off)
  • A/B test results (if any)
  • Recommendations for CRO
Frequency
Monthly.
Audience
Product team, marketing team, management.

Executive dashboard

Content

  • KPI summary (traffic, conversions, revenue, ROI)
  • Trends (up or down compared with the previous period)
  • Biggest successes and biggest risks
Frequency
Weekly or real time (dashboard).
Audience
Senior management.

How to get the most out of measurement and analytics

Focus on the right KPIs

Vanity metrics versus actionable metrics

Vanity metrics (look good but do not lead to action)

  • Total page views (say nothing about quality)
  • Social media followers (if they do not convert)
  • App downloads (if nobody uses the app)

Actionable metrics (drive decisions)

  • Conversion rate (directly linked to revenue)
  • Customer acquisition cost (CAC)
  • Customer lifetime value (LTV)
  • ROI per channel

Focus on metrics that affect the business goals.

Always compare with previous periods

An isolated figure says little
“We had 10,000 visitors this month.” Good or bad?
A comparison gives context
“We had 10,000 visitors, up from 8,000 last month, a rise of 25 per cent.” Now we can see that traffic is growing.

Compare

  • Month-over-month (MoM)
  • Year-over-year (YoY) to adjust for seasonality

Segment the data

Aggregated data hides insights.

Example

Overall conversion rate: 2 per cent

Broken down by device

  • Desktop: 3 per cent
  • Mobile: 1 per cent

Insight: The mobile experience is poor. Focus on mobile optimisation.

Segment by

  • Traffic source
  • Device
  • Geography
  • New or returning visitors

Automate reporting

Manual reports take time and are easy to forget.

Automate

  • Scheduled reports (Data Studio, email every Monday)
  • Dashboards (real-time data always available)
  • Alerts (email when traffic falls by more than 30 per cent or conversions suddenly rise)

Result: Less time spent producing reports, more time for analysis and action.

Verify the data

Data can be wrong

  • The tracking code installed several times (duplicate page views)
  • Bot traffic (artificially inflated)
  • Misconfigured filters (internal traffic is included)

Sanity check

  • Do the figures match previous periods?
  • Sudden peaks or drops? Investigate why.
  • Compare with other sources (GA versus server logs)

How data leads to better decisions

The aim of measurement and analytics is better decisions that drive growth.

The cycle

  1. Measure → Collect data
  2. Analyse → Find insights
  3. Decide → What should be done?
  4. Act → Implement actions
  5. Measure again → Did it work?

Repeat the cycle for continuous improvement.

Contact us about measurement and analytics

Carnaby helps companies set up tracking, analyse data and turn insights into concrete actions, from Google Analytics configuration to automated dashboards and data-driven marketing strategies.

Contact us →

Want to know where you stand today?

Call us, or fill in the form and we will get back to you.

Contact us Read about the pre-study