Growth or marketing activity: how to tell progress from busyness

Zeleni stol prepun marketinških aktivnosti prelazi u zlatni put održivog poslovnog rasta.

By Patricia Jarmek Marketing Strategy & Analytics 27 October 2026

6 min read

A marketing team can publish more content, grow traffic, launch new campaigns and fill a dashboard with upward arrows while the business barely changes. Activity proves that work happened. Growth means value increased—and that the increase has a reasonable chance of lasting.

Google Analytics therefore organises business-objective reporting around outcomes such as generating leads or driving sales, not around the number of campaigns produced [1]. A useful measurement system keeps operational metrics, but places them below the business outcome they are expected to influence.

Busyness is not the same as progress

Impressions, clicks, posts, visits and emails sent are execution signals. They answer, “Did we deliver the plan?” Growth asks the harder question: do we have more valuable customers, more contribution margin, better sales opportunities or a higher probability of repeat purchase?

Traffic growth without qualified demand may be more expensive noise. More leads with a lower sales-acceptance rate may slow the team down. Higher revenue with an even faster increase in acquisition cost can reduce contribution margin. The activity is real; the interpretation is wrong.

Define growth before choosing a metric

For ecommerce, growth may mean contribution after product cost, media and returns—not gross revenue alone. For a B2B company it may mean more sales-accepted opportunities and won deals, not more forms. In a subscription model, retention, account expansion and churn are central.

Write down one primary definition, the time horizon and the baseline. Then choose leading indicators. GA4 User lifetime can compare sources and campaigns by users’ lifetime value rather than only revenue in the selected month [2]. That helps separate a channel that closes fast, low-value purchases from one that acquires customers with higher long-term value.

Use a four-level measurement hierarchy

Business outcome: revenue, contribution margin, qualified pipeline, won deals or retention.
Customer behaviour: repeat purchase, activation, usage, referral or sales-stage progression.
Leading marketing signals: branded search, qualified organic traffic, target-account engagement or conversion rate.
Operational activity: spend, campaigns, content, ads and tests shipped.

Every lower level needs an explainable relationship with the level above. If it has none, keep it in an operational dashboard, but do not present it as evidence of growth.

Cohorts reveal whether growth lasts

A monthly total can rise simply because the company buys a fresh audience every month. Cohort exploration follows related groups over time and can show how long newly acquired users take to transact or whether retention improves after a product or campaign change [3].

Compare acquisition-month cohorts by second purchase, 90-day revenue or activation. In B2B, compare lead cohorts by sales acceptance, stage velocity and win rate. Growth that disappears as soon as spend falls looks more like rented volume than an asset the company has built.

An attributed result is not necessarily incremental growth

A platform may attribute a conversion to a campaign because a touchpoint occurred inside its window. That does not prove the conversion would not have happened without the campaign. Google Ads Conversion Lift compares treatment and control groups, defining incremental conversions as the difference between them [4].

For material budget decisions, combine attribution, experiments and the business source of truth. Google’s open-source Meridian MMM connects budget planning with modelling and experimental evidence [5]. Meridian GeoX provides publisher-agnostic geographic experiments and can calibrate marketing-mix models [6]. These methods are unnecessary for every post or email, but valuable when the decision can move a large share of budget.

Manage a portfolio, not the winner of the week

Classify work into three groups: activity producing an outcome now, activity building future demand and activity producing visibility only. Track efficiency and capacity for the first. Track branded search, direct demand, audience quality and later conversions for the second. Stop the third group or turn it into a defined test.

Use weekly meetings for leading indicators and execution problems. Monthly, compare business outcomes, cohorts and unit economics. Quarterly, challenge incrementality and budget allocation. This cadence avoids punishing long-term initiatives because they lack an immediate last-click result, without letting poor performance hide behind the word “brand”.

A 30-day audit checklist

Choose one primary outcome and one owner of its source data.
Align definitions of revenue, lead and customer across CRM, WooCommerce, GA4 and ad platforms.
Add cost, margin or quality next to volume.
Introduce at least one cohort comparison.
Separate attributed results from incremental results.
For every activity metric, name the decision it can change.
Remove one report line that has not changed a decision in three months.

Growth is not an absence of activity. It is activity connected to a business change, observed long enough, and tested with a method proportionate to the importance of the decision.

Frequently asked questions

What is the most important growth metric?

It depends on the business model. Contribution margin, retained revenue, customer value or qualified pipeline are often more useful than a universal marketing metric.

Are impressions and clicks useless?

No. They are valuable diagnostic and leading signals, but without a relationship to a business outcome they do not prove growth.

How long should we wait for evidence of growth?

At least one realistic purchase or sales cycle. Retention and repeat purchase usually require comparison across several monthly cohorts.

Can revenue grow while marketing becomes less efficient?

Yes. If acquisition cost, discounting or returns rise faster than revenue, contribution margin can fall while gross revenue rises.

When is an experiment necessary?

When attribution is insufficient and the decision is material, such as a large budget increase, a brand-effect question or overlapping channels.