Strategija email kampanj

Kako meriti prihodek kampanj brez dvojnega štetja

Posted on July 18, 2026 by

Če želite meriti prihodek kampanj brez dvojnega štetja, naredite štiri stvari: v svojem email orodju preklopite pripis (atribucijo) z odpiranj na klike, skrajšajte časovno okno pripisa za kampanje na 24–72 ur, nehajte seštevati prihodek, ki si ga vsak kanal na svoji nadzorni plošči pripisuje sam, in vsak mesec preverite številko emaila glede na skupni prihodek svoje trgovine. Kadar je odločitev res pomembna, izvedite kontrolno skupino (holdout) — kampanjo pošljite 90 % segmenta, 10 % zadržite in primerjajte stopnje nakupa. Ta zadnja metoda je edina, ki meri, kaj je kampanja dejansko povzročila, in ne le, poleg česa je slučajno stala. Vse ostalo v tem članku je razlaga in točne nastavitve, kajti razlika med »email je prejšnji mesec ustvaril 12.000 €« in »email je prejšnji mesec ustvaril 4.000 €« odloča, katere kampanje boste še naprej pošiljali.

Zakaj se vsaka nadzorna plošča prilizuje sama sebi

Tukaj je neprijetna aritmetika, ki jo večina lastnikov trgovin prej ali slej naredi. Meta pravi, da je prejšnji mesec pripeljal 18.000 €. Google pravi 14.000 €. Vaša email platforma pravi 9.000 €. Vaša trgovina je naredila 30.000 €. Kanali si skupaj lastijo 41.000 € zaslug za 30.000 € prodaje.

Ne lažejo, vsaj ne čisto. Vsako orodje šteje naročilo za »svoje«, če se je stranka pred nakupom v nekem trenutku dotaknila tega kanala. Stranka v torek klikne oglas na Meti, v četrtek odpre vašo kampanjo, v petek na Googlu vtipka ime vaše znamke in kupi. Tri nadzorne plošče si vsaka pripišejo celotno naročilo. Nobenega škandala ni — tako pač deluje pripis, ko vsak sodnik hkrati igra za eno od ekip.

Težava je, kaj počnete z napihnjeno številko. Vidite »kampanje so ustvarile 9.000 €« in sklenete, da kampanje nosijo trgovino, zato jih pošiljate več, presojate jih manj kritično in morda premalo vlagate v avtomatizacije, ki tiho opravljajo pravo delo. Dvojno štetje boža ego in narobe razporeja vaš čas.

Zakaj »kar zaupaj številki platforme« ni dovolj

Skušnjava je, da izberete eno nadzorno ploščo in ji verjamete. To odpove iz konkretnega razloga: privzete nastavitve vašega email orodja so običajno najbolj velikodušne, kar jih je na voljo.

Največ škode naredita dve privzeti nastavitvi:

Pripis na podlagi odpiranj. Mnoge platforme štejejo naročilo, če je stranka email znotraj okna zgolj odprla. Toda odpiranja so nezanesljiva, odkar je Apple leta 2021 z iOS 15 uvedel Mail Privacy Protection — Apple Mail vnaprej naloži slike v emailu, vključno s sledilnim pikslom, zato je velik del vaših »odpiranj« Applovih strežnikov, ne ljudi. Naročilo, »pripisano« emailu, ki ga stranka nikoli ni prebrala, je čisti šum.
Dolga okna pripisa. Petdnevno okno pomeni, da se stranka, ki je vaš torkov email prezrla, v soboto pa kupila, ker jo je prijatelj priporočil, vseeno prikaže kot prihodek kampanje. Daljše kot je okno, več običajne prodaje v ozadju vaše kampanje posrkajo v svoje zasluge.

Naložite ti dve privzeti nastavitvi in kampanja lahko »ustvari« tisoče evrov od strank, ki je zavestno nikoli niso videle.

Kje dvojno štetje dejansko nastaja

Štiri mesta, kjer pušča, približno po velikosti:

Samodejna odpiranja, šteta kot angažiranost. Že opisano zgoraj. Če vaš pripis vključuje odpiranja, predvidevajte, da je napihnjen, in ne morete zlahka reči, za koliko.
Okna, daljša od vašega nakupnega cikla. Če večina strank, ki kliknejo, kupi v enem ali dveh dneh, potem vse onkraj 72-urnega okna večinoma pobira naključja.
Dvojno lastenje med kanali. Email, Meta in Google si vsak pripišejo celotne zasluge za skupna naročila. Seštevanje njihovih nadzornih plošč bo vedno dalo številko, večjo od resničnosti.
Prekrivanje kampanj in avtomatizacij. Stranka klikne vašo kampanjo, ne kupi, dva dni pozneje opusti košarico, dobi email o opuščeni košarici in kupi. Odvisno od logike vaše platforme si naročilo lahko lastita tako kampanja kot potek (flow). Ko primerjate »prihodek kampanj« s »prihodkom avtomatizacij«, da bi se odločili, kam vložiti trud, to prekrivanje tiho popači primerjavo — tema, ki šteje, ko se odločate, kako uravnotežiti kampanje in avtomatizirane emaile.

Čistejša metoda štetja, korak za korakom

Za to ne potrebujete diplome iz analitike. Potrebujete strožje nastavitve in eno mesečno navado.

1. korak — preklopite na pripis na podlagi klikov. Klik je namerno človeško dejanje. Je najnižja poštena meja za trditev »ta email je sodeloval pri prodaji«. Naredite to najprej; gre za eno nastavitev.

2. korak — skrajšajte okno kampanje. Nastavite ga na 24–72 ur. Izberite na podlagi svojih podatkov: poglejte, kako dolgo po kliku vaše stranke običajno naročijo. Za večino trgovin, ki prodajajo izdelke pod 150 €, se nakupi, ki jih sproži kampanja, zgodijo hitro. Trgovine z dražjimi izdelki in dvotedenskimi cikli razmisleka lahko upravičijo daljša okna — a bodite tedaj pošteni, da je številka mehkejša.

3. korak — eno platformo obravnavajte kot email semafor in nikoli ne seštevajte kanalov. Prihodek emaila prihaja iz vašega email orodja, s strogimi nastavitvami zgoraj. Celotno resnico dajo poročila o naročilih vaše trgovine. Ne gradite preglednice, ki sešteva Meta + Google + email; taka preglednica vedno laže.

4. korak — izvedite mesečno preverjanje deleža. Prihodek emaila, pripisan po klikih, delite s skupnim prihodkom trgovine. Opazujte trend, ne absolutne številke. Če delež emaila v enem mesecu poskoči s 15 % na 30 %, v katerem ste spremenili le nastavitev pripisa, je skok posledica merjenja, ne uspešnosti.

5. korak — za pomembne odločitve uporabite kontrolne skupine. Ko se odločate, ali si ponavljajoča vrsta kampanje zasluži svoje mesto, razdelite segment: 90 % dobi kampanjo, naključnih 10 % ne. V naslednjem tednu med skupinama primerjajte prihodek na osebo. Razlika je resnični, povzročeni prihodek kampanje. To je edina metoda na tem seznamu, ki loči »email je dobil zasluge« od »email je opravil prodajo« — in stane vas le nekoliko manjšo pošiljko.

Ena meja, ki jo velja potegniti: ta članek govori o tem, kako ne prekomerno šteti neposrednega prihodka kampanje. Nekatere kampanje si resnično prislužijo denar, ki ga ne boste videli v nobenem 72-urnem oknu — izobraževanje in predstavitve kategorij, ki se obrestujejo šele čez tedne. Merjenje tega zapoznelega učinka je druga vaja, obravnavana v članku kako prepoznati kampanje, ki ustvarjajo prihodnje nakupe. Stroga okna za neposredne zasluge, ločena analiza za počasne gorilnike. Ne mešajte obojega.

Ponazoritveni primer

Recimo, da trgovina naredi 40.000 € na mesec. Email platforma na privzetih nastavitvah (odpiranja, petdnevno okno) poroča o 12.000 € prihodka kampanj — 30 % trgovine. Lastnik preklopi na pripis po klikih z 48-urnim oknom. Poročani prihodek kampanj pade na 4.500 €.

Nič se ni poslabšalo. Kampanje počnejo natanko to, kar so vedno. A zdaj lastnik vidi, katere pošiljke dejansko premikajo izdelke: dve od šestih kampanj v mesecu sta ustvarili skoraj vseh 4.500 €, ostale štiri pa so se peljale na valu pripisa po odpiranjih. Tiste štiri je treba premisliti na novo. (Številke so ponazoritvene — vaša razmerja bodo drugačna, a smer padca je značilna, ko zaostrite nastavitve.)

Kaj spremljati vsak mesec

Prihodek kampanj, pripisan po klikih z vašim fiksnim oknom — delovni semafor.
Prihodek na prejemnika po kampanji, da velik seznam ne prikrije šibke pošiljke.
Delež emaila v skupnem prihodku trgovine, spremljan skozi mesece ob nespremenjenih nastavitvah.
Dvig od kontrolne skupine (holdout lift) pri eni ali dveh najpomembnejših ponavljajočih vrstah kampanj, testiran nekajkrat na leto.
Stopnja odjav na kampanjo — kajti pošiljka, ki »naredi« 500 € in pri tem sežge 80 naročnikov, ni zmaga v vrednosti 500 €.

Doslednost tukaj premaga natančnost. Nekoliko nepopolna metrika, merjena vsak mesec enako, vam pove več kot popolna, ki jo nenehno na novo opredeljujete. Če pregledujete celoten lijak, ne le email, vključite to v revizijo konverzij spletne trgovine, ki bi jo morala izvesti vsaka trgovina.

Kako to narediti v Omnisendu

V svojih trgovinah uporabljam Omnisend, in najprej poštena opomba: kot vsaka email platforma je tudi njegov poročani prihodek povsem odvisen od nastavitev pripisa, ki jih izberete — orodje vam bo z veseljem pokazalo velikodušno številko, če pustite velikodušne nastavitve. Preverite konfiguracijo pripisa v svojem računu in jo nastavite na klike s kratkim oknom, namesto da sprejmete, kar je privzeto.

Kar mi je pri tem konkretnem opravilu všeč: poročila o kampanjah prikazujejo prihodek in metrike vrste prihodka za vsako pošiljko v enem pogledu, tako da je primerjava kampanj med seboj ob enakih nastavitvah hitra, gradnja 90/10 kontrolne skupine pa je zgolj logika segmentov — ustvarite segment, ga razdelite, zadržani del izključite iz pošiljke. Nobene telovadbe s preglednicami. Omnisend je partner Shopimationa v okviru partnerskega programa; priporočam ga, ker ga uporabljam vsak dan, ne ker bi z nekim posebnim orodjem vprašanje pripisa izginilo. Ne izgine.

Vaš naslednji korak

Še danes odprite nastavitve pripisa v svoji email platformi in si zapišite, kakšne trenutno so. Večina lastnikov trgovin ni nikoli pogledala. Če najdete pripis po odpiranjih ali okno, daljše od 72 ur, ga spremenite, nato pa z novimi številkami znova preberite prejšnjemesečno poročilo o kampanjah. Kampanje, ki preživijo strožje štetje, so vaše resnične zvezde — in ko enkrat veste, katere so, se splača preizkusiti, ali jih lahko izboljšate s pravim postopkom A/B-testiranja.

Email Campaign Strategy

Measuring Campaign Revenue Without Overcounting

Posted on July 17, 2026 by

To measure campaign revenue without overcounting, do four things: switch your email tool’s attribution from open-based to click-based, shorten the attribution window to 24–72 hours for campaigns, stop adding up the revenue that each channel’s dashboard claims for itself, and sanity-check email’s number against your store’s total revenue every month. When a decision really matters, run a holdout — send the campaign to 90% of the segment, keep 10% back, and compare purchase rates. That last method is the only one that measures what a campaign actually caused rather than what it happened to sit next to. Everything else in this article is the reasoning and the exact settings, because the difference between “email made €12,000 last month” and “email made €4,000 last month” changes which campaigns you keep sending.

Why every dashboard flatters itself

Here’s the uncomfortable arithmetic most store owners eventually run. Meta says it drove €18,000 last month. Google says €14,000. Your email platform says €9,000. Your store did €30,000. The channels are collectively claiming €41,000 of credit for €30,000 of sales.

They’re not lying, exactly. Each tool counts an order as “theirs” if the customer touched that channel at some point before buying. A customer clicks a Meta ad on Tuesday, opens your Thursday campaign, googles your brand name on Friday, and buys. Three dashboards each book the full order. No scandal there — that’s how attribution works when every referee also plays for a team.

The problem is what you do with the inflated number. You see “campaigns generated €9,000” and conclude campaigns are carrying the store, so you send more of them, judge them less carefully, and maybe under-invest in the automations quietly doing the real work. Overcounting flatters your ego, and it misallocates your time.

Why “just trust the platform’s number” isn’t good enough

The tempting fix is to pick one dashboard and believe it. That fails for a specific reason: your email tool’s default settings are usually the most generous ones available.

Two defaults do most of the damage:

Open-based attribution. Many platforms count an order if the customer merely opened an email within the window. But opens have been unreliable since Apple shipped Mail Privacy Protection with iOS 15 back in 2021 — Apple Mail preloads email images, including the tracking pixel, so a large share of your “opens” are Apple’s servers, not humans. An order “attributed” to an email the customer never actually read is pure noise.
Long attribution windows. A five-day window means a customer who ignored your Tuesday email but bought on Saturday because a friend recommended you still shows up as campaign revenue. The longer the window, the more ordinary background sales your campaigns absorb credit for.

Stack those two defaults and a campaign can “generate” thousands of euros from customers who never consciously saw it.

Where the overcounting actually happens

Four leak points, in rough order of size:

Auto-opens counted as engagement. Covered above. If your attribution includes opens, assume it’s inflated and you can’t easily say by how much.
Windows longer than your buying cycle. If most of your customers who click buy within a day or two, anything beyond a 72-hour window is mostly collecting coincidences.
Cross-channel double-claiming. Email, Meta, and Google each take full credit for shared orders. Summing their dashboards will always produce a number bigger than reality.
Campaign vs. automation overlap. A customer clicks your campaign, doesn’t buy, abandons a cart two days later, gets the abandoned cart email, and purchases. Depending on your platform’s logic, both the campaign and the flow may claim it. When you compare “campaign revenue” to “automation revenue” to decide where to spend effort, this overlap quietly distorts the comparison — a topic that matters when you’re deciding how to balance campaigns and automated emails.

The cleaner counting method, step by step

You don’t need an analytics degree. You need stricter settings and one monthly habit.

Step 1 — switch to click-based attribution. A click is a deliberate human action. It’s the minimum honest bar for saying “this email participated in the sale.” Do this first; it’s one setting.

Step 2 — shorten the campaign window. Set it to 24–72 hours. Pick based on your own data: look at how long after clicking your customers typically order. For most stores selling under €150 products, purchases triggered by a campaign happen fast. High-ticket stores with two-week consideration cycles can justify longer windows — but then be honest that the number is softer.

Step 3 — treat one platform as the email scoreboard and never sum channels. Email revenue comes from your email tool, with the strict settings above. Total truth comes from your store’s order reports. Don’t build a spreadsheet that adds Meta + Google + email; that spreadsheet always lies.

Step 4 — run the monthly share check. Divide click-attributed email revenue by total store revenue. Watch the trend, not the absolute number. If email’s share jumps from 15% to 30% in a month where you only changed an attribution setting, the jump is measurement, not performance.

Step 5 — use holdouts for decisions that matter. When you’re deciding whether a recurring campaign type earns its place, split the segment: 90% get the campaign, a random 10% don’t. Compare revenue per person between the groups over the following week. The difference is the campaign’s real, caused revenue. It’s the only method on this list that separates “email got the credit” from “email made the sale” — and it costs you nothing but a slightly smaller send.

One boundary worth drawing: this article is about not overcounting a campaign’s direct revenue. Some campaigns genuinely earn money you won’t see in any 72-hour window — education and category introductions that pay off weeks later. Measuring that delayed effect is a different exercise, covered in how to identify campaigns that create future purchases. Strict windows for direct credit, separate analysis for the slow burners. Don’t blend the two.

An illustrative example

Say a store does €40,000 a month. The email platform, on default settings (open-based, five-day window), reports €12,000 in campaign revenue — 30% of the store. The owner switches to click-based attribution with a 48-hour window. Reported campaign revenue drops to €4,500.

Nothing got worse. The campaigns are doing exactly what they always did. But now the owner can see which sends actually move product: two of the month’s six campaigns produced almost all of the €4,500, and the other four were riding the open-attribution wave. Those four get rethought. (Numbers illustrative — your ratios will differ, but the direction of the drop is typical when tightening settings.)

What to track monthly

Click-attributed campaign revenue with your fixed window — the working scoreboard.
Revenue per recipient per campaign, so a big list doesn’t disguise a weak send.
Email’s share of total store revenue, trended over months with settings held constant.
Holdout lift on your one or two most important recurring campaign types, tested a few times a year.
Unsubscribe rate per campaign — because a send that “makes” €500 while burning 80 subscribers is not a €500 win.

Consistency beats precision here. A slightly imperfect metric measured the same way every month tells you more than a perfect one you keep redefining. If you’re auditing your whole funnel and not just email, fold this into the ecommerce conversion audit every store should run.

Doing this in Omnisend

I run Omnisend in my own stores, and the honest note first: like every email platform, its reported revenue depends entirely on the attribution settings you choose — the tool will happily show you a generous number if you leave generous settings in place. Check your account’s attribution configuration and set it to clicks with a short window rather than accepting whatever the default is.

What I like for this specific job: campaign reports show revenue and revenue-type metrics per send in one view, so comparing campaigns against each other on equal settings is quick, and building the 90/10 holdout is just segment logic — create the segment, split it, exclude the holdout from the send. No spreadsheet gymnastics. Omnisend is an affiliate partner of Shopimation; I recommend it because it’s what I use daily, not because the attribution question disappears with any particular tool. It doesn’t.

Your next step

Open your email platform’s attribution settings today and write down what they currently are. Most store owners have never looked. If you find open-based attribution or a window longer than 72 hours, change it, then re-read last month’s campaign report with the new numbers. The campaigns that survive the stricter count are your real performers — and once you know which they are, it’s worth testing whether you can make them better with a proper A/B testing process.