The Digital Marketing and Measurement Model: How to Actually Use Google Analytics

For small business owners and marketers who have GA4 installed, check it occasionally, and still couldn't say what it actually means for the business. The tool isn't the problem — there's a strategy step missing before you ever open it.

Before you open GA4, define what the business is actually trying to achieve, break that into specific goals, pick a handful of KPIs that measure progress, set a real target for each one, and always look at the data segmented — never in aggregate. This five-step sequence is Avinash Kaushik's Digital Marketing and Measurement Model. Most GA4 dashboards skip straight to metrics and skip the four steps that would have told you which metrics matter.

---

The dashboard everyone has and nobody trusts

Almost every business with a website has a GA4 dashboard open somewhere. Sessions this month versus last month. A bounce rate nobody's quite sure is good or bad. A conversion number that goes up and down for reasons nobody can explain.

The tool isn't the problem. GA4 collects an enormous amount of accurate data. The problem is that most people open it and start looking at metrics before they've decided what the metrics are supposed to prove. Avinash Kaushik — Google's longtime Digital Marketing Evangelist and the author of Web Analytics 2.0 — built a simple model to fix exactly this failure. It's five steps, done in order, and none of them require opening an analytics tool:

Business Objective → Goals → KPIs → Targets → Segments

Most businesses jump straight to the fourth step — staring at metrics — with none of the first three done. That's why the dashboard feels like noise.

---

Step 1 — Business Objective

Start with the actual reason the website or campaign exists, in plain business language — not "increase traffic." Traffic isn't a business objective; nobody's business survives on traffic alone. A real objective sounds like "grow revenue from online sales," "generate qualified appointment requests," or "reduce the cost of acquiring a new client."

If you can't state the objective without the word "traffic," "engagement," or "awareness" doing all the work, you don't have one yet — you have a proxy for one.

Step 2 — Goals

Break the objective into a small number of specific goals that would prove it's happening. For "grow revenue from online sales," that might be: increase the number of completed purchases, and increase average order value. For "generate qualified appointment requests," it might be: increase the number of booked consultations, and improve the quality (show-up rate) of the people who book.

Goals are still not metrics yet — they're the specific outcomes a metric will eventually measure.

Step 3 — KPIs

For each goal, choose the one or two metrics that actually indicate progress toward it — not the metrics that are easiest to find in GA4. This is where most dashboards go wrong: they fill up with whatever GA4 shows by default (sessions, users, pageviews, average engagement time) instead of the specific numbers tied to the goals from Step 2.

For "increase completed purchases," the KPI is purchase conversion rate, segmented by channel — not sessions. For "increase booked consultations," it's consultation form completions per visitor, not pageviews. Every metric you didn't deliberately choose for a goal doesn't belong on the dashboard; it belongs in a deeper exploration report you check when something looks off.

Step 4 — Targets

A KPI without a number attached to it is just a number floating in space — you need a specific target and a timeframe. "Improve conversion rate" isn't a target. "Move purchase conversion rate from 1.8% to 2.5% over the next quarter" is. Without a target, there's no way to know if last month's 2.1% is good news, bad news, or noise.

Targets should come from your own historical baseline and realistic effort, not an industry benchmark pulled from a blog post — your 2.1% might be excellent for a considered, high-ticket purchase and mediocre for an impulse buy.

Step 5 — Segments

Never make a decision from an aggregated, site-wide number — always look at it broken down. This is the step Kaushik is most insistent about, to the point of a catchphrase: segment or die. A 2% overall conversion rate is meaningless on its own. It might be 8% for returning visitors on mobile and 0.3% for first-time visitors arriving from paid social — two entirely different problems hiding inside one average.

At minimum, segment every KPI by: new vs. returning visitors, channel (organic, paid, direct, referral, email), and device. If a KPI moves and you can't say which segment moved it, you don't understand what happened yet.

---

Putting all five steps together

Step	Example A — e-commerce	Example B — service business
Objective	Grow revenue from online sales	Generate more qualified consultations
Goal	Increase completed purchases and order value	Increase booked, qualified consultations
KPI	Purchase conversion rate; average order value	Consultation bookings per 100 visitors
Target	1.8% → 2.5% conversion rate this quarter	15 → 25 bookings per 1,000 visitors this quarter
Segment	By channel, new vs. returning, device	By channel, landing page, traffic source

Notice that "sessions" and "pageviews" never appear as a KPI in either example. They can still be useful context — but they don't belong at the top of the dashboard, because neither business objective is actually about traffic volume.

---

Macro conversions vs. micro conversions

Once the five steps are set, one more Kaushik concept makes the model work in practice: most visitors will not complete your macro conversion on any given visit — so you also need to track the smaller steps that show they're moving toward it.

Type	Definition	Examples
Macro conversion	The primary outcome the business exists to produce	A completed purchase, a signed contract, a booked appointment
Micro conversion	A smaller action that signals progress toward the macro conversion	Adding to cart, viewing the pricing page, starting (not finishing) a form, spending several minutes on a service page, returning for a second visit

If you only measure the macro conversion, a page that gets people 80% of the way there and a page nobody reads past the headline look identical in your dashboard — both show zero. Micro conversions are what let you tell them apart, and they're usually the first thing to fix when a macro conversion rate is stuck.

---

Where this goes wrong in practice

The most common mistake isn't picking the wrong KPI — it's skipping straight to metrics with none of the first three steps done. A few patterns to watch for:

• The dashboard has 20+ metrics and no stated goal for most of them. If a number doesn't trace back to a goal from Step 2, it's clutter, not insight.
• A KPI exists with no target. "We're tracking form completions" isn't a plan; "we're moving form completions from 40 to 60 a month" is.
• A conclusion gets drawn from an unsegmented number. "Conversion rate is down" said about the whole site, when it's actually down for one channel and flat everywhere else, sends the team fixing the wrong thing.
• Vanity metrics get treated as KPIs. Followers, pageviews, and session count feel good to report but rarely map to a real business objective — see our Views Aren't Growth post for the content-specific version of this same problem.

---

Setting this up in GA4

The model is deliberately tool-agnostic — you do all five steps before opening GA4 — but here's where each step lands once you do:

1. KPIs become GA4 conversion events. Mark your macro conversion (purchase, form submit, booking) and your key micro conversions as conversions in GA4's Events configuration, not just as events you happen to collect.
2. Targets live outside GA4. GA4 won't show you a target line — track it in whatever dashboard or spreadsheet your team actually reviews, next to the real number.
3. Segments come from Explorations, not the default reports. Build a saved Exploration that breaks your key conversion events down by channel, device, and new-vs-returning by default, so segmentation is the normal view, not an extra step someone has to remember to add.
4. Channel data needs to be trustworthy first. If AI-referred traffic is showing up as "Direct" in your channel grouping, your segments will be wrong before you even start — see our GA4 attribution fix for AI traffic if that applies to you.

---

The model, in one line

Don't ask "what does the data say?" until you've answered "what were we trying to prove, and how would we know?" Objective, Goals, KPIs, Targets, Segments — in that order, every time. GA4 is good at answering questions. It has never been good at telling you which questions to ask.

Once the objectives and goals are set, the next question is which specific metrics actually earn a spot on the KPI list — see How to Choose the Right Marketing Metrics and KPIs for the filtering logic.

Ready to put a real measurement model behind your marketing?

JYNLAB helps small and mid-sized businesses build the analytics setup that actually answers "is this working" — not just a dashboard of numbers going up and down. If you want a partner to help define the objectives, KPIs, and GA4 setup behind your marketing, book a demo call and we'll walk through where you stand.

Book a Demo Call