Nizka prodaja kljub obisku

Kako najti strani, na katerih odnehajo obiskovalci z močnim nakupnim namenom

Posted on July 18, 2026 by

Da bi našli strani, na katerih odnehajo obiskovalci z močnim nakupnim namenom, nehajte gledati skupni delež zavrnitev (bounce rate) in začnite slediti ljudem, ki so že pokazali nakupne signale — tistim, ki so iskali, dodali v košarico ali prišli do koraka z dostavo. Pripravite lijakovo poročilo, ki pokaže, koliko sej preide vsako stopnjo (ogled izdelka → dodajanje v košarico → košarica → dostava → plačilo → naročilo), poiščite en sam največji odstotni padec med dvema korakoma in nato oglejte deset ali petnajst posnetkov sej ljudi, ki so odpadli prav tam. Stran z najstrmejšim kvalificiranim osipom — in ne tista z največ izhodi na splošno — je tam, kjer vam pušča denar. Ta vodnik razloži, kako to naredite brez podatkovne ekipe.

Zakaj je stran z »največ izhodi« napačna za popravljanje

Vsaka trgovina ima stran z veliko izhodi. Običajno je to domača stran ali objava na blogu, ker tam pristane hladen, na pol radoveden promet in nato odide. Popravljanje te strani se zdi produktivno, a spremeni skoraj nič, saj večina teh ljudi tako ali tako nikoli ne bi kupila.

Obiskovalci z močnim nakupnim namenom so drugačna skupina. V iskalnik so vtipkali ime točno določenega izdelka, nekaj so dodali v košarico, kliknili so »na blagajno«. Ta dejanja so zaveze. Ko nekdo, ki je v košarico dal izdelek za 70 €, izgine na strani z dostavo, to ni brskalec, ki je odtaval drugam — to je skorajšnji kupec, ki je zadel ob zid. Eden takih je vreden dvajset zavrnitev na domači strani. Vprašanje torej ni »katera stran izgubi največ ljudi«, temveč »katera stran izgubi največ ljudi, ki so bili tik pred plačilom«.

Pravi problem: vaš lijak skrije puščanje znotraj povprečja

Večina lastnikov gleda eno samo številko — recimo 2,1-odstotno stopnjo konverzije — in celotno trgovino obravnava kot problem. To povprečje zlije skupaj blagajno, ki deluje odlično, in stran izdelka, ki tiho ubija polovico svojega kvalificiranega prometa. Povprečja ne morete popraviti. Popravite lahko le korak.

Tega vam skupna stopnja konverzije nikoli ne pove: osip skoraj nikoli ni enakomerno porazdeljen. V večini trgovin, ki sem jih pregledal, resnično škodo naredita en ali dva prehoda, ostali pa so zdravi. Morda vaše strani izdelkov spodobno pretvarjajo v košarico, prehod iz košarice na blagajno je v redu, nato pa 60 % ljudi izgine med stranjo z dostavo in stranjo s plačilom. Ta zadnji prehod je vaše puščanje. Vse pred njim deluje. Če vlijete več prometa, samo potisnete več ljudi proti istemu pokvarjenemu koraku — plačate obisk in ga izgubite vsakič na istem mestu.

Kje se izguba dejansko zgodi (in kako jo oceniti)

Preden se česa lotite, dajte puščanju številko, da veste, ali je vredno vašega popoldneva. Iz analitike potrebujete le tri podatke: seje, ki so dosegle korak, seje, ki so dosegle naslednji korak, in vašo povprečno vrednost naročila.

Preprost ponazoritveni izračun:

Recimo, da 4.000 sej na mesec doseže stran z dostavo na blagajni.
Le 1.600 jih doseže stran s plačilom — 60-odstotni padec na tem enem prehodu.
Vaše povprečno naročilo znaša 65 €.

Če bi se že četrtina teh 2.400 izgubljenih sej pretvorila ob popravljenem koraku z dostavo, je to 600 naročil × 65 € = 39.000 € na mesec, ki tičijo v eni sami strani. (Ponazoritveno — vstavite svoje številke, a izračun naredite, preden se odločite, da ni vredno popravka.)

Ta številka je bistvo. »Izboljšaj blagajno« je nejasna zadolžitev. »Med dvema klikoma tiči približno 39.000 € na mesec« je razlog, da si očistite koledar.

Praktičen način, kako najti stran, korak za korakom

Ne potrebujete specialista. Potrebujete pa, da naredite te korake po vrsti.

1. Sestavite ali odprite lijakovo poročilo

V GA4 uporabite raziskovanje lijaka (Funnel exploration), v Shopify vam analiza blagajne pokaže košarica → blagajna → nakup, večina orodij za snemanje sej prav tako gradi lijake. Korake opredelite kot resnično pot, ki jo kupec ubere v vaši trgovini. Glejte odstotek, ki se premakne z vsakega koraka na naslednjega, ne surovih številk. Največji padec na enem koraku je vaš glavni osumljenec.

2. Segmentirajte samo na visok nakupni namen

Ponovno poženite ta lijak za ljudi, ki so pokazali namen: seje z dodajanjem v košarico, seje iz iskanja po blagovni znamki ali vrnjeni obiskovalci. Če korak, ki je za vse skupaj videti povprečen, za to skupino izpade grozljivo, ste našli, kje prav zavezani kupci odnehajo. To je ostrejši signal kot zmešana številka. Če niste prepričani, da so vaši signali namena zanesljivi, sedem signalov, da je obiskovalec zainteresiran, a še ne prepričan razčleni, katera vedenja dejansko napovedujejo nakup.

3. Oglejte si osip, ne ugibajte ga

Zdaj odprite deset do petnajst posnetkov sej ljudi, ki so odpadli na vašem najslabšem koraku. To je del, ki ga večina lastnikov preskoči, in prav tu se navadno skriva odgovor. Vzorec boste hitro opazili: besni kliki na polje za kupon, ki noče delovati, strošek dostave, ki se pojavi pozno in pošlje kazalec naravnost na gumb za zapiranje, gumb »Plačaj« na mobilnem telefonu, potisnjen pod rob zaslona, obrazec, ki zavrne veljavno telefonsko številko. Petnajst posnetkov premaga uro teoretiziranja.

4. Preberite signale vedenja na strani

Na sumljivo stran dodajte toplotne zemljevide (heatmap) ali zemljevide drsenja. Če ljudje nikoli ne zdrsnejo do gumba za dodajanje v košarico, je stran predolga ali pa je pasica zavajajoča. Če se ustavijo nad dostavo ali vračili in odidejo, je težava vaša politika, ne vaš izdelek. Podatki o namenu izhoda (exit-intent) na tej eni strani vam povedo, kaj so gledali v trenutku, ko so odnehali.

Kaj avtomatizirati, ko poznate stran

Iskanje puščanja je polovica dela. Druga polovica je ujeti ljudi, ki zdrsnejo mimo, medtem ko stran popravljate — kajti tudi popolna stran izgubi nekaj kvalificiranih kupcev zaradi raztresenosti, preverjanja cen ali jokajočega malčka.

Nastavite tok za obnovitev, sprožen z vedenjem, usmerjen točno na korak, kjer so odnehali:

Sprožilec: dosegel korak z visokim osipom (dodal v košarico, začel blagajno, izdelek si ogledal trikrat) in odšel brez naročila.
Segment: samo visok namen — ima e-poštni naslov, pokazal je zavezujoče dejanje. Ne obstreljujte hladnih brskalcev.
Časovnica: prvo sporočilo v eni uri, dokler je namen še topel, drugo naslednji dan, tretje po dveh do treh dneh, če izdelek ali marža to opravičujeta.
Kanal: najprej e-pošta, za košarice višje vrednosti dodajte SMS, če imate privolitev.
Vsebina: točno tisti izdelek, ki so ga pustili, jasen odgovor na ugovor, ki ste ga videli v posnetkih (strošek dostave, vračila, velikosti), in en jasen gumb nazaj na ta korak. Brez mreže desetih izdelkov.
Cilj: obnoviti naročilo in vam obenem po tihem povedati, kateri ugovor ustavi največ ljudi.

Tok za obnovitev in popravek strani delujeta skupaj. Nadaljevanje, ki zapre vrzel med nekomerkim zanimanjem in nakupom, si zasluži svojo pozornost — vrzel v nadaljevanju med zanimanjem za izdelek in nakupom opisuje, kako ga sestaviti, da ljudi zgolj ne nadleguje.

Kako izmeriti, ali ste popravili pravo stran

Po vsaki spremembi spremljajte tri stvari:

Konverzijo iz koraka v korak na popravljenem prehodu. To je neposredna semaforska plošča. Če padec z dostave na plačilo pade s 60 % na 40 %, ste premaknili iglo tam, kjer šteje.
Obnovljena naročila iz sproženega toka. Prihodek, ki bi ga sicer izgubili, zdaj ujet, medtem ko izpopolnjujete stran.
Skupno stopnjo konverzije — preverjeno nazadnje. Naj se dvigne, a je to zaostajajoča, zmešana številka. Najprej zaupajte metriki koraka, povprečje to potrdi pozneje.

Spreminjajte eno stvar naenkrat. Popravite polje za kupon, počakajte na dovolj sej, preberite nov lijak. Če v enem popoldnevu prepišete stran z dostavo, predelate politiko vračil in dodate značke zaupanja, nikoli ne boste vedeli, kaj je delovalo — in če se številka poslabša, ne boste vedeli, kaj razveljaviti.

Kje se vpne Omnisend

Iskanje strani je analitično delo, to zmore vsako lijakovo orodje. Kjer si orodje prisluži svoje mesto, je tok za obnovitev, ki ujame kvalificirane kupce v trenutku, ko zdrsnejo. V svojih trgovinah uporabljam Omnisend — preizkusil sem ga v primerjavi s Klaviyom in ostal zaradi bolj samodejne nastavitve ter združene e-pošte, SMS-a in potisnih sporočil v enem toku. Sprožiti sporočilo ob »začel blagajno, a ni dokončal«, potegniti nazaj točno tisti izdelek in dodati SMS le za košarice nad določenim pragom je vzelo popoldne, ne razvijalca.

Poštene omejitve: avtomatizacija obnovi ljudi, ki iztečejo mimo strani, ne popravi pa strani. Če je težava vaš strošek dostave, je noben e-poštni niz ne prehiti — najprej popravite korak, nato pustite toku, da pobere ostalo. Omnisend je partner v pridruženem programu Shopimation, priporočam ga iz vsakodnevne uporabe; brezplačni paket je dovolj, da na najslabšem koraku sestavite en tok za obnovitev in vidite, ali se izplača.

Vaš naslednji korak

Ta teden odprite eno lijakovo poročilo, ga segmentirajte na seje z dodajanjem v košarico in poiščite svoj en sam najstrmejši padec. Ta ena številka vam pove, katero stran naj naslednjič spremljate prek posnetkov. Ko veste, kje odnehajo obiskovalci z močnim nakupnim namenom, je naravno nadaljevanje razumeti, zakaj sploh brskajo, a se nikoli ne zavežejo — zakaj obiskovalci brskajo po vaši trgovini, a nikoli ne pridejo do blagajne — in ko boste pripravljeni celotno stvar sistematizirati, revizija konverzije v spletni trgovini, ki bi jo morala izvesti vsaka trgovina ta enkratni lov na eni strani spremeni v ponovljiv pregled.

Low Sales Despite Traffic

How to Find the Pages Where High-Intent Shoppers Give Up

Posted on July 18, 2026 by

To find the pages where high-intent shoppers give up, stop looking at your overall bounce rate and start following the people who already showed buying signals — the ones who searched, added to cart, or reached the shipping step. Pull a funnel report that shows how many sessions pass each stage (product view → add to cart → cart → shipping → payment → order), find the single biggest percentage drop between two steps, then watch ten or fifteen session recordings of people who fell out there. The page with the steepest qualified drop-off, not the most exits overall, is where your money is leaking. This guide walks through how to do that without a data team.

Why “most exits” is the wrong page to fix

Every store has a page with a lot of exits. Usually it’s the homepage or a blog post, because that’s where cold, half-curious traffic lands and leaves. Fixing that page feels productive and changes almost nothing, because most of those people were never going to buy.

High-intent shoppers are a different group. They typed a specific product name into search, they added something to the cart, they clicked “checkout.” Those actions are commitments. When someone who added a €70 item to their cart disappears at the shipping page, that’s not a browser wandering off — that’s a near-customer hitting a wall. One of those is worth twenty homepage bounces. So the question isn’t “which page loses the most people,” it’s “which page loses the most people who were about to pay.”

The real problem: your funnel hides the leak inside an average

Most owners look at one number — a 2.1% conversion rate, say — and treat the whole store as the problem. That average smears together a checkout that works fine and a product page that quietly kills half its qualified traffic. You can’t fix an average. You can only fix a step.

Here’s the thing an aggregate conversion rate never tells you: the drop is almost never spread evenly. In most stores I’ve looked at, one or two transitions do the real damage while the rest are healthy. Maybe your product pages convert to cart at a decent clip, your cart-to-checkout is fine, and then 60% of people vanish between the shipping page and the payment page. That last transition is your leak. Everything upstream is working. Pour more traffic in and you just push more people toward the same broken step — you pay for the visit and lose it at the same spot every time.

Where the loss actually happens (and how to size it)

Before you touch anything, put a number on the leak so you know it’s worth your afternoon. You only need three figures from your analytics: sessions that reached a step, sessions that reached the next step, and your average order value.

A simple illustrative calculation:

Say 4,000 sessions a month reach the checkout’s shipping page.
Only 1,600 reach the payment page — a 60% drop at that single transition.
Your average order is €65.

If even a quarter of those 2,400 lost sessions would have converted with a fixed shipping step, that’s 600 orders × €65 = €39,000 a month sitting in one page. (Illustrative — plug in your own numbers, but do the math before you decide it’s not worth fixing.)

That figure is the point. “Improve the checkout” is a vague chore. “There’s roughly €39k a month stuck between two clicks” is a reason to clear your calendar.

The practical way to find the page, step by step

You don’t need a specialist. You need to do these in order.

1. Build or open a funnel report

In GA4, use the Funnel exploration; in Shopify, the checkout analysis shows cart → checkout → purchase; most session-recording tools build funnels too. Define the steps as the real path a buyer takes in your store. Look at the percentage moving from each step to the next, not the raw counts. The biggest single-step drop is your prime suspect.

2. Segment to high-intent only

Re-run that funnel for people who showed intent: sessions with an add-to-cart, sessions from branded search, or returning visitors. If a step that looked mediocre for everyone looks terrible for this group, you’ve found where committed buyers specifically give up. That’s a sharper signal than the blended number. If you’re not sure your intent signals are trustworthy, the seven signals that a visitor is interested but not yet convinced breaks down which behaviors actually predict buying.

3. Watch the fallout, don’t guess it

Now open ten to fifteen session recordings of people who dropped at your worst step. This is the part most owners skip, and it’s where the answer usually lives. You’ll see the pattern fast: rage-clicks on a coupon field that won’t apply, a shipping cost that appears late and sends the cursor straight to the close button, a mobile “Pay” button pushed below the fold, a form that rejects a valid phone number. Fifteen recordings beat an hour of theorizing.

4. Read the on-page behavior signals

Add heatmaps or scroll maps on the suspect page. If people never scroll to the add-to-cart button, the page is too long or the hero is misleading. If they hover on shipping or returns and leave, your policy is the problem, not your product. Exit-intent data on that one page tells you what they were looking at the moment they quit.

What to automate once you know the page

Finding the leak is half the job. The other half is catching the people who slip through while you fix it — because even a perfect page loses some qualified shoppers to distraction, price-checking, or a crying toddler.

Set a behavior-triggered recovery flow aimed at the exact step where they gave up:

Trigger: reached the high-drop step (added to cart, started checkout, viewed a product three times) and left without ordering.
Segment: high-intent only — has an email, showed the commitment action. Don’t blast cold browsers.
Timing: first message within an hour while intent is warm, a second the next day, a third after two to three days if the item or margin justifies it.
Channel: email first; add SMS for higher-value carts if you have consent.
Content: the specific product they left, a plain answer to the objection you saw in the recordings (shipping cost, returns, sizing), and one clear button back to that step. No ten-product grid.
Goal: recover the order and, quietly, tell you which objection stops the most people.

The recovery flow and the page fix work together. The follow-up that closes the gap between someone’s interest and their purchase deserves its own attention — the follow-up gap between product interest and purchase covers how to build it so it doesn’t just nag people.

How to measure whether you fixed the right page

Watch three things after each change:

Step-to-step conversion at the fixed transition. This is the direct scoreboard. If the shipping-to-payment drop goes from 60% to 40%, you moved the needle where it counts.
Recovered orders from the triggered flow. Revenue you’d have lost outright, now captured while you iterate on the page.
Overall conversion rate — checked last. It should rise, but it’s a lagging, blended number. Trust the step metric first; the average confirms it later.

Change one thing at a time. Fix the coupon field, wait for enough sessions, read the new funnel. If you rewrite the shipping page, redo the returns policy, and add trust badges all in one afternoon, you’ll never know which one worked — and if the number gets worse, you won’t know what to undo.

Where Omnisend fits

Finding the page is analytics work; any funnel tool does it. Where a tool earns its place is the recovery flow that catches qualified shoppers the moment they slip. I use Omnisend in my own stores — I tested it against Klaviyo and stayed for the more hands-off setup and the combined email, SMS, and push in one flow. Triggering a message off “started checkout but didn’t finish,” pulling the exact product back in, and adding an SMS only for carts over a threshold took an afternoon, not a developer.

Honest limits: automation recovers the people who leak past a page; it does not fix the page. If your shipping cost is the problem, no email sequence outruns that — fix the step first, then let the flow mop up the rest. Omnisend is an affiliate partner of Shopimation, and I recommend it from daily use; the free tier is enough to build one recovery flow on your worst step and see whether it pays.

Your next step

Open one funnel report this week, segment it to add-to-cart sessions, and find your single steepest drop. That one number tells you which page to watch recordings of next. Once you know where high-intent shoppers give up, the natural follow-on is understanding why they browse but never commit in the first place — why visitors browse your store but never reach the checkout — and, when you’re ready to systematize the whole thing, the ecommerce conversion audit every store should run turns this one-page hunt into a repeatable check.