What marketing analytics actually means
Marketing analytics is the practice of collecting, measuring, and interpreting data about your marketing to understand what's working and what isn't. That's it. Everything else — dashboards, attribution models, conversion tracking, channel reporting — is in service of that single goal.
The reason it matters: marketing without measurement is just spending money and hoping. With measurement, every campaign either teaches you something useful (even if it fails) or scales something that works.
My background is in finance before marketing. Moving from a world where every number is auditable to one where most marketing teams track vanity metrics was a significant culture shock. The discipline of treating marketing data the way a finance team treats a P&L is the single biggest mindset shift that moves marketing from a cost centre to a growth engine.
Why most businesses measure the wrong things
The most common mistake in marketing analytics: optimising for metrics that feel good but don't connect to revenue. Follower count. Impressions. Page views. Email list size. These are proxy metrics — they can correlate with business outcomes, but they are not business outcomes.
The useful question is always: what happens after this number goes up? If follower count increases by 1,000, does revenue increase? Does pipeline increase? Does anything measurable change? If the honest answer is "we don't know," then follower count is a vanity metric and you're not measuring the right thing.
The metrics that actually matter
Here's the framework I use for prioritising what to track, organised by funnel stage:
Awareness metrics
- Organic search impressions and clicks (Google Search Console) — are new people finding you through search?
- Branded vs non-branded search split — are people searching for your name, or for what you do? Branded searches are warm; non-branded show you're reaching new audiences.
- Social reach — only matters if you can connect it to the next stage. Reach without conversion is noise.
Consideration metrics
- Email list growth rate — week-on-week, are you adding subscribers? Is the rate accelerating or decelerating?
- Resource download completions — are people consuming your lead magnets all the way through?
- Session-to-lead rate — of all the people who visit your site, what percentage give you their email? This is the most actionable MOFU metric.
Conversion metrics
- Contact form submissions or Calendly bookings — are people taking the final action?
- Cost per lead (if you're running paid) — how much does it cost to generate a qualified enquiry?
- Lead-to-close rate — of all enquiries, what percentage become clients? This tells you about lead quality, not just quantity.
The tool stack
The minimum viable analytics stack for a marketing consultant or small B2B business:
- Google Analytics 4 (GA4): the foundation. Tracks all on-site behaviour — page views, sessions, events, conversions. Free. Set up conversion events for form submissions, Calendly clicks, and resource downloads.
- Google Search Console: tracks your organic search performance — keywords you're ranking for, click-through rates, indexing status. Free. Essential for SEO work.
- Looker Studio: connects to GA4, Search Console, and your social platforms to build a single reporting dashboard. Free. I build a weekly review dashboard in here that takes 10 minutes to review.
- Google Tag Manager: manages all your tracking tags without touching code. Lets you add GA4 events, conversion pixels, and third-party scripts without a developer for most tasks.
For more advanced setups, Plausible (privacy-first, lighter than GA4) or PostHog (product analytics, great for understanding in-product behaviour) are worth adding once the basics are solid.
How to build a useful weekly report
The weekly review is where analytics becomes useful. Here's the format I use — it takes about 15 minutes:
- Organic traffic this week vs last week — is it up or down? By how much?
- Top 5 pages by traffic — what content is pulling the most visitors? Is it changing week-on-week?
- Conversion count — how many form submissions / Calendly bookings / resource downloads this week?
- Email list change — how many new subscribers this week? Which source drove them?
- One actionable insight — what does this week's data suggest doing differently next week?
The last point is the only one that matters. Data without interpretation is just numbers. The job of a weekly report is to surface one decision worth making.
Turning data into decisions
Here's the process I run when a number moves unexpectedly:
- Isolate the change: is this a real change or statistical noise? A week-on-week swing of 10% is usually noise. A persistent trend over 3+ weeks is real.
- Find the cause: did something change in what you published, promoted, or linked to? Did a competitor do something? Did a third-party platform change its algorithm?
- Form a hypothesis: "I think this is happening because X. If I'm right, doing Y should move Z."
- Test the hypothesis: make one change. Wait two weeks. Measure the effect.
- Document what you learned. The insight from this cycle is more valuable than any single metric.