Rastoči stroški pridobivanja kupcev

Zakaj lahko tudi vaša najboljša oglaševalska kampanja pridobi nedonosne kupce

Posted on July 18, 2026 by

Zato ker naslovne številke kampanje – donos na porabljeni oglaševalski denar, strošek na nakup, stopnja pretvorbe – vse merijo prvo naročilo, ne kupca. Vaša najboljša kampanja lahko doseže odličen ROAS tako, da privabi ljudi, ki kupijo enkrat, s popustom, in se nikoli ne vrnejo. Preštejte strošek blaga, dostavo, vračila in spodbudo, ki jih je pritegnila, in del teh »zmag« je v minusu. »Najboljša kampanja« in »donosni kupci« se ocenjujeta na dveh različnih semaforjih, oglaševalska platforma pa vam pokaže le enega.

Kampanja je torej lahko vaš najboljši izvajalec in vaš največji vir kupcev, ki izgubljajo denar, hkrati. Takole se to zgodi in kako to ujeti, preden razširite izgube.

Pravi problem: razširili ste zmagovalko, dobiček pa ni sledil

Našli ste kampanjo, ki je delovala. ROAS je presegel vaš cilj, strošek na nakup je bil nizek, zato ste storili razumno stvar in vanjo vložili več denarja. Poraba je narasla. Naročila so narasla. Skupni dobiček pa je ostal enak ali zdrsnil, na način, ki ga poročilo kampanje ne zna pojasniti.

Prav ta razkorak – jasno zmagovalna kampanja v paru z razočaranjem pri dobičku – je eden najpogostejših in najdražjih vzorcev v plačanem e-trgovanju. Drag je prav zato, ker je kampanja videti dobra, zato je nagon, da vanjo naložimo več. Če pridobiva nedonosne kupce, več proračuna hitreje kupi več njih. Sorodna različica tega na ravni računa je zakaj lahko višja oglaševalska poraba prinese manj dejanske rasti; ta članek govori o tem, kako to počne posamezna, na videz močna kampanja.

Zakaj vas zmagovalne meritve lahko zavedejo

ROAS in strošek na nakup nista napačna. Sta le ozka. Štiri stvari zdrsnejo mimo njiju.

ROAS je prihodek, ne marža. ROAS 3,0 pri 35-odstotni marži pusti le tanek prispevek, ko odštejete strošek blaga, dostavo in provizije – nikakor ne potrojenega denarja. Kampanja lahko preseže vaš cilj ROAS in vseeno izgublja denar na naročilo.
Šteje eno naročilo. Meritev se konča pri prvem nakupu. Ali se ta kupec sploh kdaj vrne – tisto, kar naredi pridobivanje donosno – ji je nevidno.
Kreativa izbira kupca. Oglas, ki ga vodi popust, zanesljivo privabi kupce, ki jih vodi popust. Ti se dobro pretvorijo (odlične številke pri prvem naročilu) in slabo ponavljajo (grozne življenjske številke). Kampanja optimizira prav k tistim kupcem, ki se ne bodo vrnili. Ta učinek izbire je srce zakaj oglasi, ki temeljijo na popustih, pogosto poslabšajo ekonomiko pridobivanja kupcev.
Vračila in pripisovanje stvari še bolj zameglijo. Nekatere kampanje kopičijo vračila; nekatere kampanje ponovnega ciljanja si pripisujejo zasluge za naročila, ki bi se zgodila tako ali tako. Oboje polepša poročilo in izčrpava banko.

Kje se zgodi izguba – znotraj »zmagovalnega« naročila (ponazoritveno)

Vzemite kampanjo, ki beleži ROAS 3,0. Sliši se kot obdržalka. Spremljajte eno naročilo skozi. Številke so ponazoritvene.

Oglas oglašuje 30 % popusta. Kupec kupi košaro za 60 €, torej s popustom poberete 42 €. Strošek blaga v tej košari je 24 €. Dostava, ki jo krijete, 6 €. Provizije za plačilo in platformo 2 €. Prispevek pred stroškom pridobitve: 42 € − 24 € − 6 € − 2 € = 10 €.

Zdaj strošek pridobitve. Strošek te kampanje na nakup je 14 €. Torej ste pri tem naročilu 4 € v rdečih številkah – pri kampanji, ki poroča ROAS 3,0. Meritev na osnovi prihodka je rekla zmaga. Marža je rekla izguba. In ker je kreativa s 30 % popusta privabila lovca za ugodnostmi, so možnosti za drugo naročilo po polni ceni, ki bi rešilo račun, majhne. Razširite to kampanjo in razširili boste izgubo 4 € preko tisočev naročil, medtem ko nadzorna plošča še naprej sveti zeleno.

Edino, kar bi jo lahko obrnilo v plus, je ponovni nakup. Ali pride, je vprašanje kohorte, ne vprašanje poročila kampanje – in to je številka, ki je večina trgovin nikoli ne izvleče. Celotno težo kupca, ki se nikoli ne vrne, najdete v skriti strošek pridobivanja kupcev, ki kupijo le enkrat.

Praktična rešitev: sodite kampanje po kupcih, ki jih pustijo za seboj

Spremenite, kako ocenjujete kampanjo, po vrstnem redu vpliva.

Ocenjujte po prispevku, ne po prihodkovnem ROAS. Ponovno izračunajte donos vsake kampanje z uporabo marže po strošku blaga, dostavi, provizijah in njenem lastnem popustu. Nekatere zmagovalke ob tej luči tiho preidejo v negativno. To so tiste, ki jih je treba postaviti pod vprašaj, ne razširiti.
Označite kupce po kampanji, ki jih je pridobila. Nato spremljajte kohorto. Kupci katere kampanje znova naročijo v 90 dneh? Kateri ustvarijo vračila? Kampanja s skromnim ROAS in močno stopnjo ponovnih nakupov lahko premaga »najboljšo« kampanjo z odličnim ROAS in brez ponavljanja.
Vsaki odločitvi o širjenju dodajte preverjanje stopnje ponovnih nakupov. Preden povečate proračun, poglejte 90-dnevno stopnjo ponovnih nakupov kupcev, ki jih je ta kampanja že pripeljala. Dober ROAS ob slabem ponavljanju je znak za stop, ne zelena luč.
Popravite spodbudo, če je kreativa problem. Če se kampanja pretvarja le z globokim popustom, verjetno pridobiva napačne kupce. Preizkusite pristope, ki jih vodi vrednost, in privabljajo ljudi, ki hočejo izdelek, ne le cene.

Kaj avtomatizirati: dajte majavi kampanji priložnost, da se povrne

Ni vsak kupec, pridobljen s popustom, izgubljen primer. Nekatere je mogoče naknadno spremeniti v ponovne kupce – in tam avtomatizacija spremeni izid mejne kampanje.

Sprožilec: prvi nakup.
Segment: prvi kupci, označeni po pridobitveni kampanji, da vidite, katere kohorte se odzovejo.
Časovnica: uvajanje in izobraževanje o izdelku v prvih dneh; poziv k ponovnemu naročilu ali navzkrižni prodaji na naravnem ciklu izdelka; ponovna pridobitev, če utihnejo.
Kanal: e-pošta za vsebino, SMS za časovno občutljive spodbude k ponovnemu naročilu, kjer imate soglasje.
Vsebina: razlogi za vrnitev po polni ceni – iztržite več iz nakupa, naravni naslednji izdelek – namesto novega popusta, ki znova utrjuje navado lova za ugodnostmi.
Cilj: dvigniti stopnjo ponovnih nakupov kupcev, ki jih je kampanja pridobila, kar je edino, kar lahko kampanjo na pragu rentabilnosti premakne v dejanski dobiček.

Če kampanja ostane nedonosna tudi potem, ko tokovi opravijo svoje, je to vaš dokaz, da jo odrežete – čisto, s kohortnimi podatki za odločitvijo.

Kaj meriti

ROAS na osnovi prispevka na kampanjo – marža, ne prihodek. Ena najbolj razjasnjujoča sprememba, ki jo lahko naredite.
90-dnevna stopnja ponovnih nakupov po pridobitveni kampanji – loči prave zmagovalke od laskavih.
Stopnja vračil po kampanji – nekatere zmagovalke skrijejo svoje izgube v čakalni vrsti za vračila.
Strošek na donosnega kupca, ne na nakup – nakup ni dobiček; ta meritev ohranja razliko na očeh.

Prav ta zadnja razlika – obisk in naročila v primerjavi z resnično donosnimi kupci – leži pod celotnim tem problemom in jo je vredno prebrati v celoti: razlika med poceni obiskom in donosnimi kupci.

Kje se v to vključi Omnisend

Oglaševalska platforma vam pove, da je kampanja zmagala. Ali so kupci, ki jih je pridobila, vredni obdržati, se odloči po prodaji, v tokovih, ki jim jih pošljete. V svojih lastnih trgovinah uporabljam Omnisend – izbral sem ga pred Klaviyem, potem ko sem zagnal oba – ker to poprodajno delo naredi praktično: označite in segmentirajte kupce po tem, od kod so prišli, zaženite ponakupne tokove in tokove za ponovno pridobitev, da preizkusite, ali je mogoče kupce kampanje spremeniti v ponovne, in preberite prihodek na prejemnika po segmentu, da je odgovor v podatkih, ne v slutnji.

Iskrena meja: avtomatizacija lahko reši nekatere kupce, ki jih je pripeljala šibka kampanja. Ne more pa popraviti kampanje, ki strukturno pridobiva ljudi, ki nikoli ne bodo plačali polne cene. Ko kohortni podatki pravijo, da se kupci kampanje ne vračajo niti po dobrem spremljanju, je prava poteza spremeniti kampanjo, ne pošiljati več e-pošte. Omnisend je partner Shopimationa v pridruženem programu; priporočam ga iz resnične uporabe, brezplačni paket pa zadostuje, da označite kohorte in preizkusite ponakupni tok.

Vaš naslednji korak

Izberite kampanjo, na katero ste najbolj ponosni, in njen donos ponovno sestavite na dva načina: enkrat na prihodkovnem ROAS, enkrat na prispevni marži po njenem lastnem popustu, nato preverite 90-dnevno stopnjo ponovnih nakupov kupcev, ki jih je pripeljala. Če je prispevek tanek in ponavljanje nizko, ste našli »zmagovalko«, ki jo je vredno postaviti pod vprašaj, preden jo razširite. Da vidite isto past s strani obiska, nato preberite zakaj lahko višja oglaševalska poraba prinese manj dejanske rasti.

Rising Customer Acquisition Costs

Why Your Best Ad Campaign Can Still Create Unprofitable Customers

Posted on July 18, 2026 by

Because a campaign’s headline numbers — return on ad spend, cost per purchase, conversion rate — all measure the first order, not the customer. Your best campaign can hit a great ROAS by attracting people who buy once, at a discount, and never come back. Count the cost of goods, shipping, returns, and the incentive that pulled them in, and a chunk of those “wins” are underwater. “Best campaign” and “profitable customers” are scored on two different boards, and the ad platform only shows you one of them.

So a campaign can be your top performer and your biggest source of money-losing buyers at the same time. Here’s how that happens, and how to catch it before you scale the losses.

The real problem: you scaled the winner and profit didn’t follow

You found a campaign that worked. The ROAS beat your target, cost per purchase was low, so you did the sensible thing and put more money behind it. Spend went up. Orders went up. And overall profit stayed flat, or slid, in a way the campaign report can’t explain.

That disconnect — a clearly winning campaign paired with disappointing profit — is one of the most common and most expensive patterns in paid ecommerce. It’s expensive precisely because the campaign looks good, so the instinct is to feed it more. If it’s recruiting unprofitable customers, more budget buys more of them faster. The related version of this at the account level is why higher ad spend can produce less real growth; this article is about how a single strong-looking campaign does it.

Why the winning metrics can lie to you

ROAS and cost per purchase aren’t wrong. They’re just narrow. Four things slip past them.

ROAS is revenue, not margin. A 3.0 ROAS on 35% margins leaves only a slim contribution once COGS, shipping, and fees come out — nothing like triple your money. A campaign can clear your ROAS target and still lose money per order.
It counts one order. The metric ends at the first purchase. Whether that buyer ever returns — the thing that makes acquisition pay — is invisible to it.
The creative selects the customer. A discount-led ad reliably attracts discount-led buyers. They convert well (great first-order numbers) and repeat poorly (terrible lifetime numbers). The campaign optimizes toward exactly the customers who won’t come back. That selection effect is the heart of why discount-driven ads often make acquisition economics worse.
Returns and attribution muddy it further. Some campaigns concentrate returns; some retargeting campaigns claim credit for orders that would have happened anyway. Both flatter the report and drain the bank.

Where the loss happens — inside a “winning” order (illustrative)

Take a campaign posting a 3.0 ROAS. Sounds like a keeper. Follow one order through. Illustrative numbers.

The ad promotes 30% off. A customer buys a €60 basket, so with the discount you collect €42. COGS on that basket is €24. Shipping you cover is €6. Payment and platform fees, €2. Contribution before acquisition cost: €42 − €24 − €6 − €2 = €10.

Now the acquisition cost. This campaign’s cost per purchase is €14. So on that order you’re €4 in the red — on a campaign reporting a 3.0 ROAS. The revenue-based metric said win. The margin said loss. And because the 30%-off creative attracted a deal-seeker, the odds of a full-price second order that rescues the math are low. Scale this campaign and you scale the €4 loss across thousands of orders while the dashboard keeps showing green.

The one thing that could turn it positive is a repeat purchase. Whether it comes is a cohort question, not a campaign-report question — and it’s the number most stores never pull. The full weight of a buyer who never returns is in the hidden acquisition cost of customers who buy only once.

The practical fix: judge campaigns by the customers they leave behind

Change how you grade a campaign, in order of impact.

Grade on contribution, not revenue ROAS. Rebuild each campaign’s return using margin after COGS, shipping, fees, and its own discount. Some winners quietly go negative under this light. Those are the ones to question, not scale.
Tag customers by the campaign that acquired them. Then follow the cohort. Which campaign’s buyers reorder within 90 days? Which produce the returns? A campaign with a modest ROAS and a strong repeat rate can beat a “top” campaign with a great ROAS and no repeat.
Add a repeat-rate check to every scale decision. Before you increase budget, look at the 90-day repeat rate of the customers that campaign already brought in. Good ROAS plus poor repeat is a stop sign, not a green light.
Fix the incentive if the creative is the problem. If the campaign only converts with a deep discount, it’s likely recruiting the wrong buyers. Test value-led angles that attract people who want the product, not just the price.

What to automate: give a shaky campaign a chance to pay back

Not every discount-acquired customer is a lost cause. Some can be converted into repeat buyers after the fact — and that’s where automation changes the outcome of a marginal campaign.

Trigger: first purchase.
Segment: first-time buyers, tagged by acquisition campaign so you can see which cohorts respond.
Timing: onboarding and product education in the first days; a reorder or cross-sell prompt on the product’s natural cycle; a win-back if they go quiet.
Channel: email for the substance, SMS for time-sensitive reorder nudges where you have consent.
Content: reasons to come back at full price — get more from the purchase, the natural next product — rather than another discount that re-teaches the deal-seeking habit.
Goal: lift the repeat rate of customers a campaign acquired, which is the only thing that can move a break-even campaign into real profit.

If a campaign stays unprofitable even after the flows do their work, that’s your evidence to cut it — cleanly, with cohort data behind the call.

What to measure

Contribution-based ROAS per campaign — margin, not revenue. The single most clarifying change you can make.
90-day repeat rate by acquisition campaign — separates real winners from flattering ones.
Return rate by campaign — some winners hide their losses in the returns queue.
Cost per profitable customer, not per purchase — a purchase isn’t a profit; this metric keeps the difference in view.

That last distinction — traffic and orders versus genuinely profitable customers — sits underneath this whole problem, and it’s worth reading in full: the difference between cheap traffic and profitable customers.

Where Omnisend fits

The ad platform tells you a campaign won. Whether the customers it won are worth keeping is decided after the sale, in the flows you send them. I use Omnisend in my own stores — I chose it over Klaviyo after running both — because it makes that after-sale work practical: tag and segment buyers by where they came from, run post-purchase and win-back flows to test whether a campaign’s customers can be turned into repeaters, and read revenue per recipient by segment so the answer is in the data, not a hunch.

The honest boundary: automation can rescue some customers a weak campaign brought in. It can’t fix a campaign that structurally recruits people who will never pay full price. When the cohort data says a campaign’s buyers don’t come back even after good follow-up, the right move is to change the campaign, not send more email. Omnisend is an affiliate partner of Shopimation; I recommend it from real use, and the free tier is enough to tag cohorts and test a post-purchase flow.

Your next step

Pick the campaign you’re most proud of and rebuild its return two ways: once on revenue ROAS, once on contribution margin after its own discount, then check the 90-day repeat rate of the customers it brought in. If contribution is thin and repeat is low, you’ve found a “winner” worth questioning before you scale it. To see the same trap from the traffic side, read why higher ad spend can produce less real growth next.