One event – one shot per user. It tells you where to invest, which campaigns are burning budget, and where to raise the alarm. Worth checking regularly.

When first_visit beats sessions hands down

In GA4 sessions are everywhere – in dashboards, reports, campaigns. But sessions measure activity, not acquisition. One engaged customer visiting the store daily is 30 sessions a month. Two new customers who came and left are 2 sessions. Which source is really working toward growth?

first_visit fires exactly once per device, on the user’s first visit. It can’t be duplicated. It won’t be inflated by returning customers, remarketing or bots browsing the catalog. That makes it the only metric in GA4 that answers one specific question: how many new people reached my store, and from where?

Sessions measure activity. first_visit measures whether your store really attracts new people.

The difference shows best in concrete cases:

SituationSessionsfirst_visit
A regular customer comes back 5 times a month+5 sessions – inflates traffic+0 – already counted
A remarketing campaign reactivates 200 people+200 sessions – looks great+0 – these aren’t new customers
A new Instagram post draws 80 people+80 sessions+80 first_visit – real acquisition
PMAX delivers 500 clicks a day+500 sessions – but how many are new?e.g. +120 first_visit – that many are actually new
A QR code on packaging is scanned by 50 customers+50 sessions – like any other+50 first_visit + the highest purchase intent

That’s why a first_visit analysis gives a completely different perspective than a standard session report – especially when you run campaigns across several channels at once and want to know what’s really working for new customers.

Where to invest? The channels that are growing

The first question in a first_visit analysis is always the same: where did the new users come from and what’s changing? More interesting than the ranking itself is which channels are growing – that’s where unnoticed potential may lie.

Where do new users come from? DataOrganizer MCP
GA4 · 90 days
Show me the first_visit breakdown by channel for the last 90 days. I want to know what’s growing – not just the ranking.
AI
get_traffic_sources · event=first_visit · days=90 · compare=prev_90
Fetching the data…

207,835 first_visit in the last 90 days – an overall result of –17.1% vs the previous 90 days. Channel breakdown:

Organic Search67,637 · 32.5% ▼ –18.7%
Paid Social41,363 · 19.9% ▼ –24.3%
Direct27,655 · 13.3% ▼ –10.5%
Paid Search23,478 · 11.3% ▼ –11.7%
Cross-network (PMAX)22,776 · 11.0% ▼ –18.1%
Paid Shopping5,182 · 2.5% ▲ +181%
Offline / QR on packaging3,363 · 1.6% ▲ +8.3%
Instagram organic1,514 · 0.7% ▲ +25.7%

Three channels are growing despite the overall trend: Paid Shopping (+181%), Instagram organic (+26%) and QR codes on packaging (+8%). All the rest are losing.

Paid Shopping
+181%
Previously overlooked – today it delivers 5,000+ new users a quarter and keeps growing. A classic underinvested channel.
QR on packaging
3,363
Users with the product in hand. The highest purchase intent of all first_visit channels.
Instagram organic
+26%
Growing with no media budget. A signal that organic creatives work – just increase the frequency.

Which creatives to rework?

The channel ranking is the intro. The real value lies deeper – at the campaign level. Here you can see which creatives stopped attracting new users. The key rule: look at the change, not the volume. A campaign can be #1 in absolute terms, but if its share is falling month over month – something isn’t working.

Which campaigns are losing? DataOrganizer MCP
GA4 · campaigns
Compare campaigns by first_visit – May 1–24 vs April 1–24. What’s losing, what’s growing?
AI
get_marketing_performance · event=first_visit · 2026-05-01:05-24 vs 2026-04-01:04-24
Comparing the periods…

71,428 first_visit (May 1–24) vs 84,576 (Apr 1–24) → –15.5%. Changes per campaign:

CampaignMayAprΔ
ColourLife – New customers (Meta)8242,168–62%
TOF PMAX – New customers6,9099,613–28%
PMAX – Core assortment2,6793,412–21%
Brand Search3,9124,228–7%
MOF PMAX – Brand-aware4,4664,360+2%
PMAX – Organizers2,1341,339+59%
Brand cat. bedroom2,1981,816+21%
Blog – interior inspirations751469+60%

Takeaway: the entire –15.5% is largely one campaign – ColourLife – New customers lost –62%. This isn’t a problem with the Meta channel, it’s a problem with that specific creative or audience. Before cutting the budget it’s worth checking ad fatigue and testing a new variant.

The trap of channel thinking

It’s easy to say “Meta doesn’t work” and cut the budget. But the –45% in Paid Social came from one campaign (–62%). The rest of the channel worked normally. first_visit shows this at the campaign level – you don’t lose months to a wrong diagnosis.

When to raise the alarm

Four signals in the first_visit data that should make you pause a little longer.

One channel drops 40%+ M/M
Before you cut the budget – check the campaigns. Big channel drops often come from one campaign, not a strategic failure of the whole channel.
Organic Search falls 2+ months in a row
It’s the most important acquisition channel with no media cost. A steady decline is a signal to audit SEO – before you lose positions that are hard to rebuild.
Direct traffic above 20%
A high “Direct” in first_visit often means untagged traffic – missing UTMs or campaign tagging errors. Part of your money is invisible in the reports.
Newsletter loses fast (–30%+ Y/Y)
Email is a channel that’s slow to build and quick to fall. A sharp drop in first_visit from email is a signal – check deliverability, open rate and the freshness of your list.

How often to check first_visit

You don’t need to do it daily. But there are rhythms that work well.

MonthlyRecommended
An M/M comparison on the same window – the basic rhythm. 15 minutes with an LLM and GA4 data is enough to see whether something is changing across channels.
After a new campaignAlways
A new channel, a new budget, a new creative – check first_visit after 2 weeks. It’s a faster signal than waiting for purchases.
QuarterlyStrategically
A full 90-day comparison with the previous quarter. Look for long trends in channels – things you can’t see in a monthly window.
On a sudden dropImmediately
Any sudden cliff in the trend is a signal. Don’t wait for the report – check right away which campaign or channel is behind the drop.

The main takeaway

first_visit is worth checking regularly. Here’s why.

It can’t be duplicated. It fires once per user – and that’s why it’s the cleanest acquisition metric in GA4. Sessions lie because of returning visitors. first_visit tells the truth about whether your store really attracts new people.

It shows you where to invest (Paid Shopping is up +181% while completely underinvested), which creatives to rework (one Meta campaign accounted for –62% and dragged the whole channel down) and where to raise the alarm (Organic Search falling a second month in a row, Direct above 20% – a tagging problem). You make the decision.

When you’re running acquisition campaigns – check daily. One question is enough not to miss an activity that’s just starting to pay off. First-visit ROAS will always be weaker than other activities, but that’s exactly why it’s worth keeping an eye on: a good idea can go unnoticed without regular tracking.


DataOrganizer · MCP

Ask your data these questions

DataOrganizer MCP connects an LLM with your GA4 and sales – without exporting, without tables in a spreadsheet.

DataOrganizer

Turn store data into decisions.

Stop guessing. Connect all your data sources, store, Google Analytics, Facebook Ads and more, and start acting on real numbers.

Get started free

© 2026 Datadiary · Made in Poland · KRS 0001017418 · NIP 9721336108