SMS marketing

Kako določiti, kateri kupci naj prejemajo SMS-e

Posted on July 18, 2026 by

Ni nujno, da vsi, ki so se prijavili, prejmejo vsak SMS. Odločitveno pravilo, ki deluje: kupec naj prejme SMS, ko so hkrati resnične tri stvari — dal je izrecno privolitev, sporočilo je zanj zares nujno ali osebno, in pričakovana vrednost pošiljanja presega strošek pošiljanja, pri čemer strošek pomeni tako provizijo na sporočilo kot tveganje odjave. Email je vaš privzeti kanal; SMS je izjema, ki jo prihranite za trenutke in ljudi, kjer je prekinitev upravičena. V praksi to pomeni kupce pred brskalci, dejavne naročnike pred tihimi in trenutke visoke namere pred splošnimi obvestili. Ta članek govori o tem, kdo naj prejema SMS-e. Kako zgraditi segmente same, pokriva kako segmentirati SMS-naročnike po nakupnem vedenju, kdaj pošiljati, pa je spet ločeno vprašanje.

Težava: “pošlji vsem” je privzeta izbira in je draga

Vsaka SMS-platforma naredi pošiljanje celotnemu seznamu za pot najmanjšega odpora. Eno občinstvo, eno sporočilo, en gumb. In za razliko od emaila, kjer dodatnih tisoč pošiljanj stane praktično nič, ima vsak SMS pripeto ceno na sporočilo. Pošljite kampanjo 8.000 naročnikom, ko jih je le 2.000 imelo kakršno koli realno možnost za nakup, in ste plačali 6.000 sporočil, ki so povzročila prekinitve namesto naročil.

Račun je viden strošek. Nevidni je hujši. SMS-odjave so brutalne in trajne — nekdo odgovori STOP in ta kanal je zanj izgubljen, običajno za vedno. Vsak nerelevanten SMS, ki ga pošljete napačni osebi, porabi malo tistega zaupanja, zaradi katerega je sploh delila svojo številko. Email-odnos lahko obnovite prek zavihka s promocijami. Na telefonu ni zavihka s promocijami.

Zakaj “prijavili so se, torej je v redu” ne drži

Privolitev je tla, ne odločitev. Nekdo, ki je pred osmimi meseci obkljukal polje ob zaključku nakupa, je pristal, da bo slišal od vas; ni pristal, da bo slišal vse od vas. Prijava vam pove, da je pošiljanje zakonito. Ne pove ničesar o tem, ali je to pošiljanje, tej osebi, danes, pametno.

Druga mikavna bližnjica — prekopirati email-občinstvo v SMS — pade iz istega razloga. Kanala imata drugačno ekonomiko in drugačno raven tolerance. Email, ki ga 90 % prejemnikov prezre, vas stane skoraj nič. SMS, ki ga 90 % prejemnikov prezre, je stal pravi denar in razjezil 90 % ljudi, s katerimi ste si najbolj želeli odnosa na ravni telefona. Če gre sporočilo na oba kanala, tudi to zahteva namensko obravnavo — kako se izogniti pošiljanju istega sporočila po emailu in SMS-u obravnava prav to prekrivanje.

Odločitveni okvir treh filtrov

Vsak načrtovan SMS spustite skozi tri filtre, po vrsti. Pošiljanje mora prestati vse tri.

Filter 1: Privolitev in svežina

Izrecna SMS-prijava, pravilno zbrana in ne prastara. Številka, zbrana pred dvema letoma, ki ji nikoli niste poslali SMS-a, je bolj tvegana, kot je videti: oseba je morda pozabila, da se je prijavila, in pozabljena privolitev se obnaša kot brez privolitve, ko pridejo odgovori STOP in prijave neželene pošte. Če sedite na starem, nikoli nagovorjenem seznamu, ga previdno ogrejte ali ponovno potrdite, namesto da bi ga obravnavali kot pripravljeno zalogo za pošiljanje.

Filter 2: Moč odnosa

Svoj prijavljeni seznam grobo razvrstite takole, od najbolj SMS-vrednih do najmanj:

Ponovni kupci — kupili so vsaj dvakrat, poznajo vašo znamko in dobro odmerjen SMS berejo kot storitev.
Nedavni prvi kupci — eno naročilo, še vroči. Dobri kandidati za SMS-e v zvezi z naročilom in obnovo zaloge, previdni kandidati za promocije.
Dejavni nekupci — prijavljeni prek pojavnega obrazca, klikajo vaša sporočila, niso kupili. Vredni SMS-a za močne ponudbe, vredni opazovanja glede osipa.
Tihi naročniki — prijavljeni, nikoli kliknili, nikoli kupili. Pošiljanje tej skupini je tam, kjer živi večina zapravljene porabe. Nagovarjajte jih redko, z najboljšim gradivom, ali pa sploh ne.

Ni vam treba imeti deset segmentov že prvi dan. Že ločitev “je kupil” od “ni kupil” spremeni ekonomiko vsake kampanje. Mehanika gradnje teh skupin iz podatkov o naročilih spada v kako segmentirati SMS-naročnike po nakupnem vedenju.

Filter 3: Ujemanje sporočila z osebo

Zadnji filter je za vsako kampanjo posebej: ali to konkretno sporočilo zdaj zares pomeni tej konkretni skupini? Obvestilo o obnovljeni zalogi pomeni ljudem, ki so zanj prosili, in nikomur drugemu. Bliskovita razprodaja pomeni ljudem, ki kupujejo ob promocijah. SMS z VIP-zgodnjim dostopom pomeni vašim največjim kupcem, prvemu kupcu razprodajnega izdelka pa zveni kot šum. Nujno in osebno si prisluži prekinitev. Zgolj informativno gre na email.

Kje pušča denar: ilustrativni izračun

Recimo, da vas SMS-i stanejo približno pet centov vsak in ima vaš seznam 10.000 prijavljenih številk, od katerih jih je 3.000 kdaj kupilo. (Ilustrativne številke povsod — vstavite svoje stopnje.)

Promocijsko pošiljanje celotnemu seznamu stane okoli 500 €. Če skoraj vsa nastala naročila pridejo iz segmenta kupcev plus tanek rezec dejavnih nekupcev — kar je tipična oblika —, potem je grobo 300–350 € te porabe šlo v tiho večino za skoraj ničelen donos, in to preden štejemo odjave, ki jih je sprožila prav med naročniki, ki bi jih najbolj želeli obdržati. Ciljajte namesto tega na 3.000 kupcev plus, recimo, 1.000 dejavnih nekupcev: račun pade na okoli 200 €, prihodek se komaj premakne, prihodek na sporočilo — številka, ki vam pove, ali SMS deluje — pa se grobo podvoji. Širšo sliko stroškov, vključno s krediti in provizijami platforme, si je vredno v celoti ogledati v izračun resničnega stroška SMS-marketinga.

Kaj avtomatizirati, da se odločitev sprejme sama

Bistvo okvira je, da ga enkrat zakodirate, ne da bi ga pred vsako kampanjo znova pretresali.

Sprožilec: zgradite filtre kot shranjene segmente, ki se posodabljajo samodejno — “s SMS-privolitvijo IN kupil v zadnjih 180 dneh”, “s SMS-privolitvijo IN nikoli kupil IN kliknil v zadnjih 60 dneh” in tako naprej.
Segment: vsako promocijsko kampanjo privzeto usmerite na segment kupcev in dejavnih. Pošiljanje širše od tega postane izjema, ki jo mora nekdo zavestno izbrati.
Časovnica: transakcijski in tokovni SMS-i (težave z dostavo, obvestila o obnovljeni zalogi, obnova košarice) ciljajo sami po sprožilcu; kampanje so tam, kjer živi disciplina občinstva.
Logika kanala: če je kontakt zelo dejaven na emailu, premislite, ali za to sporočilo SMS sploh potrebuje — prihranite telefon za tisto, česar email ne zmore dovolj hitro.
Cilj: vsako pošiljanje gre ljudem z verjetnim razlogom za dejanje, tihi segment pa neha tiho izčrpavati vaš proračun.

Še ena skupina si zasluži SMS-e, ki jih večina trgovin pozabi: kupci sredi zapleta z naročilom — zamujen paket, zaplet pri plačilu, nadomestek zaradi razprodane zaloge. Ta sporočila gradijo odnos, namesto da bi ga trošila, in so pokrita v uporaba SMS-a za dragocene trenutke storitve za stranke.

Kako izmeriti, ali je vaše ciljanje pravilno

Štiri številke, preverjene po segmentu in ne po povprečju kampanje:

Prihodek na poslano sporočilo — najboljša posamezna metrika ciljanja. Rast pomeni, da vaši filtri delujejo.
Stopnja odjav na kampanjo, po segmentu — če segment nekupcev osipa pri večkratniku segmenta kupcev, zaostrite filter 2. Vztrajne težave z osipom so lastna tema: kako zmanjšati SMS-odjave v spletni trgovini.
Stopnja klikov po segmentu — pove vam, katere skupine še vedno vidijo vaše SMS-e kot vredne odpiranja.
Strošek na naročilo iz SMS-a — primerjajte ga s stroškom na naročilo iz plačanih oglasov. SMS naj udobno zmaga; če ne, pošiljate preširoko.

Kako to početi v Omnisendu

Za to v svojih trgovinah uporabljam Omnisend — preizkusil sem Klaviyo in Omnisend drug ob drugem in ostal pri Omnisendu zaradi cene in dejstva, da email in SMS živita na enem profilu kontakta, kar je natanko to, kar segmentno ciljanje SMS-a potrebuje. Segmenti, zgrajeni na nakupni zgodovini, dejavnosti in statusu privolitve, se posodabljajo sami, kampanje lahko usmerite na segment enako preprosto kot na celoten seznam, poročila pa prikazujejo prihodek na kampanjo, tako da zgornji izračun ciljanja neha biti teoretičen.

Česar orodje ne bo naredilo: odločilo vaših pragov. Ali “dejaven” pomeni 30 dni ali 90 in ali tihi naročniki dobijo eno zadnjo kampanjo ali nič, so presoje o vaši marži in vaših kupcih. Programska oprema le uveljavi, karkoli se odločite.

Vaš naslednji korak

Odprite svojo SMS-platformo in svoj prijavljeni seznam še danes razdelite na dva segmenta: je kupil, ni kupil. Naslednjo kampanjo pošljite samo kupcem in primerjajte prihodek na sporočilo s svojim zadnjim pošiljanjem celotnemu seznamu. Ta ena primerjava običajno reši spor. Nato pojdite globlje v opredelitve segmentov z kako segmentirati SMS-naročnike po nakupnem vedenju.

SMS Marketing

How to Decide Which Customers Should Receive SMS

Posted on July 17, 2026 by

Not everyone who opted in should get every text. The decision rule that works: a customer should receive an SMS when three things are true at once — they gave explicit consent, the message is genuinely urgent or personal to them, and the expected value of sending beats the cost of sending, where cost means both the per-message fee and the risk of an unsubscribe. Email is your default channel; SMS is the exception you reserve for the moments and people where interruption is justified. In practice that means buyers over browsers, engaged subscribers over silent ones, and high-intent moments over general announcements. This article is about who should get texts. How to build the segments themselves is covered in how to segment SMS subscribers by purchase behavior, and when to send is a separate question again.

The problem: “send to all” is the default, and it’s expensive

Every SMS platform makes blasting the full list the path of least resistance. One audience, one message, one button. And unlike email, where an extra thousand sends costs effectively nothing, every text has a per-message price attached. Send a campaign to 8,000 subscribers when only 2,000 had any realistic chance of buying, and you paid for 6,000 messages that produced interruptions instead of orders.

The bill is the visible cost. The invisible one is worse. SMS unsubscribes are blunt and permanent — someone replies STOP and that channel is gone for them, usually for good. Every low-relevance text you send to the wrong person spends a little of the trust that made them share their number in the first place. You can rebuild an email relationship through the promotions tab. There is no promotions tab on a phone.

Why “they opted in, so it’s fine” doesn’t hold

Consent is the floor, not the decision. Someone who ticked a box at checkout eight months ago agreed to hear from you; they did not agree to hear everything from you. Opt-in tells you sending is legal. It says nothing about whether this send, to this person, today, is smart.

The other tempting shortcut — copy the email audience over to SMS — fails for the same reason. The two channels have different economics and different tolerance levels. An email that 90% of recipients ignore costs you almost nothing. A text that 90% of recipients ignore cost real money and annoyed 90% of the people you most wanted a phone-level relationship with. If a message is going to both channels, that needs deliberate handling too — how to avoid sending the same message by email and SMS deals with that overlap directly.

A three-filter decision framework

Run every planned SMS through three filters, in order. A send has to pass all three.

Filter 1: Consent and recency

Explicit SMS opt-in, collected properly, and not ancient. A number collected two years ago that has never been texted is riskier than it looks: the person may have forgotten opting in, and forgotten consent behaves like no consent when the STOP replies and spam reports come in. If you’re sitting on an old, never-messaged list, warm it carefully or re-confirm it rather than treating it as ready-to-blast inventory.

Filter 2: Relationship strength

Rank your opted-in list roughly like this, from most SMS-worthy to least:

Repeat customers — they’ve bought at least twice, they know your brand, and a well-timed text reads as service.
Recent first-time buyers — one order, still warm. Good candidates for order-related and replenishment texts, cautious candidates for promotions.
Engaged non-buyers — opted in via a popup, click your messages, haven’t purchased. Worth texting for strong offers, worth watching for churn.
Silent subscribers — opted in, never clicked, never bought. Texting this group is where most wasted spend lives. Message them rarely, with your best material, or not at all.

You don’t need ten segments on day one. Even splitting “has purchased” from “has not” changes the economics of every campaign. The mechanics of building these groups from order data belong to how to segment SMS subscribers by purchase behavior.

Filter 3: Message-to-person fit

The last filter is per-campaign: does this specific message matter to this specific group right now? A restock alert matters to the people who asked for it and no one else. A flash sale matters to people who buy on promotions. A VIP early-access text matters to your top spenders and reads as noise to a one-time buyer of a clearance item. Urgent and personal earns the interruption. Merely informational goes to email.

Where the money leaks: an illustrative calculation

Say texts cost you around five cents each and your list is 10,000 opted-in numbers, of which 3,000 have ever bought. (Illustrative numbers throughout — swap in your own rates.)

A full-list promo blast costs about €500. If nearly all resulting orders come from the buyer segment plus a thin slice of engaged non-buyers — which is the typical shape — then roughly €300–350 of that spend went to the silent majority for near-zero return, and that’s before counting the unsubscribes it triggered among exactly the subscribers you’d most want to keep. Target the 3,000 buyers plus, say, 1,000 engaged non-buyers instead: the bill drops to about €200, revenue barely moves, and revenue per message — the number that tells you whether SMS is working — roughly doubles. The wider cost picture, including credits and platform fees, is worth a full look in calculating the real cost of SMS marketing.

What to automate so the decision makes itself

The point of the framework is to encode it once, not to re-argue it before every campaign.

Trigger: build the filters as saved segments that update automatically — “SMS-consented AND purchased in the last 180 days,” “SMS-consented AND never purchased AND clicked in the last 60 days,” and so on.
Segment: default every promotional campaign to the buyer and engaged segments. Sending wider than that becomes the exception someone has to consciously choose.
Timing: transactional and flow-based texts (shipping problems, restock alerts, cart recovery) target themselves by trigger; campaigns are where the audience discipline lives.
Channel logic: if a contact is highly engaged on email, consider whether they need the SMS at all for this message — reserve the phone for what email can’t do fast enough.
Goal: every send goes to people with a plausible reason to act, and the silent segment stops quietly draining your budget.

One more group deserves texts that most stores forget: customers in mid-order trouble — a delayed package, a payment hiccup, an out-of-stock substitution. Those messages build the relationship rather than spending it, and they’re covered in using SMS for high-value customer service moments.

How to measure whether your targeting is right

Four numbers, checked per segment rather than per campaign average:

Revenue per message sent — the single best targeting metric. Rising means your filters are working.
Unsubscribe rate per campaign, by segment — if the non-buyer segment churns at multiples of the buyer segment, tighten filter 2. Persistent churn problems are their own topic: how to reduce SMS unsubscribes in ecommerce.
Click rate by segment — tells you which groups still find your texts worth opening.
Cost per order from SMS — compare it against your paid-ads cost per order. SMS should win comfortably; if it doesn’t, you’re sending too wide.

Doing this in Omnisend

I use Omnisend for this in my own stores — I tested Klaviyo and Omnisend side by side and stayed with Omnisend for the price and the fact that email and SMS live on one contact profile, which is exactly what segment-based SMS targeting needs. Segments built on purchase history, engagement, and consent status update on their own, campaigns can be pointed at a segment as easily as at the full list, and the reporting shows revenue per campaign so the targeting math above stops being theoretical.

What the tool won’t do: decide your thresholds. Whether “engaged” means 30 days or 90, and whether silent subscribers get one last campaign or zero, are judgment calls about your margin and your customers. The software just enforces whatever you decide.

Your next step

Open your SMS platform and split your opted-in list into two segments today: has purchased, has not. Send your next campaign to buyers only and compare revenue per message against your last full-list send. That one comparison usually settles the argument. Then go deeper on the segment definitions with how to segment SMS subscribers by purchase behavior.