Od prvega do drugega nakupa

Kako segmentirati kupce z največjo verjetnostjo ponovnega nakupa

Posted on July 18, 2026 by

Kupci, ki bodo najverjetneje kupili znova, so le redko tisti, ki so ob prvem naročilu porabili največ. So tisti, katerih prvi nakup in zgodnje vedenje se ujemata z vzorcem, ki so ga vaši obstoječi ponovni kupci pokazali v isti fazi. Da jih najdete, sestavite segment iz štirih signalov, ki jih že imate: kaj so kupili, koliko pozornosti so posvetili po naročilu (odpiranja, kliki, obiski strani), kako ste jih pridobili in kdaj so nazadnje naročili. Te signale ocenite, najboljši del združite v skupino in tej skupini pošljite drugačno, zgodnejše in bolj osebno sporočilo za drugi nakup kot vsem ostalim. Ta stran govori o tem, kako zgraditi ta segment — o logiki in poljih — ne o tem, kateri izdelek ponuditi naslednjič ali kdaj poslati sporočilo.

Zakaj je “vsi, ki so enkrat kupili” napačen segment

Večina trgovin obravnava prve kupce kot eno samo skupino. En sam potek po nakupu, ena časovnica, ena ponudba, poslana vsem, ki so prestopili mejo iz obiskovalca v kupca. Zdi se pošteno. Prav to pa je razlog, da tak potek dosega slabše rezultate.

Seznam prvih naročil je mešanica ljudi s povsem različnimi možnostmi, da se vrnejo. Nekdo, ki je kupil potrošni izdelek, ki mu bo v petih tednih pošel, je odprl vsak e-mail od takrat in je prišel prek organskega iskanja imena vaše znamke, je povsem drugačna priložnost kot nekdo, ki je med črnopetkovsko akcijo pograbil enkratno darilo, je prišel prek oglasa na Meti in od takrat ni odprl nobenega sporočila. Če obema pošljete enak poziv k drugemu nakupu, bodisi z zvestim nagovarjanjem prisilite drugega, da se odjavi, bodisi prvega dolgočasite s ponudbo, ki prispe prepozno in zahteva premalo.

Namen segmentiranja je tukaj ozek: ločite kupce, ki se že nagibajo k drugemu naročilu, od tistih, ki potrebujejo več časa — ali ki se nikoli ne bi vrnili. Te skupine obravnavate drugače, ker so drugačne.

Signali, ki dejansko napovedujejo drugo naročilo

Za to ne potrebujete ekipe podatkovnih znanstvenikov. Štirje signali opravijo večino dela in vsaka e-trgovinska platforma jih beleži.

1. Kaj so kupili prvič. Izbira izdelka je najmočnejši zgodnji pokazatelj, saj različni prvi izdelki nosijo različne možnosti ponovnega nakupa. Izdelek za dopolnjevanje (kava, prehranska dopolnila, nega kože, hrana za hišne ljubljenčke) ima vgrajeno naravno uro za ponovno naročilo. Vstopni izdelek ali “poskusna velikost” nakazuje nekoga, ki vas preizkuša. Močno znižan razprodajni izdelek pogosto nakazuje lovca na ugodne cene, ki bo počakal na naslednje znižanje. Razvrstite svoj katalog po tem, kateri prvi izdelki so v preteklosti vodili do drugega naročila — že samo ta razvrstitev preoblikuje vaš segment. Poglobljena analiza te razvrstitve je projekt zase, opisan v članku kako ugotoviti, kateri prvi izdelki ustvarijo najbolj zveste kupce.

2. Angažiranost po nakupu. So odprli e-mail o pošiljanju, kliknili na vodnik “kako uporabljati”, se vrnili na stran, pustili oceno? Angažiranost v dveh tednih po naročilu je živ signal zanimanja. Kupec, ki klika skozi vaše e-maile, vam sporoča, da še vedno pozorno spremlja. Kupec, ki je utihnil, ni nujno izgubljen, ni pa pripravljen na prošnjo.

3. Vir pridobitve. Od kod je kupec prišel, vpliva na to, kako verjetno je, da se vrne — včasih bolj kot to, koliko je porabil. Iskanje po blagovni znamki in promet iz priporočil običajno prinašata stabilnejše ponovne kupce kot hladni kliki na oglase, gnani s popusti. Če naročila označite z virom, lahko segment ustrezno uravnotežite — razmislek za to razliko je vreden samostojnega razumevanja v članku zakaj nekateri kanali pridobivanja prinesejo več drugih nakupov kot drugi.

4. Nedavnost in naravni cikel izdelka. Kupec, ki je pred 25 dnevi kupil 30-dnevno zalogo, je v povsem drugačnem trenutku kot tisti, ki je isto stvar kupil včeraj. Nedavnost nekaj pomeni le glede na dejansko okno ponovnega naročila izdelka, zato jo berite v kontekstu in ne kot pavšalno “število dni od naročila”.

Kako signale spremeniti v uporaben segment

Tukaj je praktična izvedba. Ne potrebujete popolnega točkovanja — groba, poštena različica premaga dovršeno, ki je nikoli ne dokončate.

Začnite tako, da vsakemu kupcu dodelite preprost seštevek točk:

Kupil prvi izdelek z visoko stopnjo ponovnega nakupa: +2
Odprl ali kliknil vsaj en e-mail po nakupu: +1
Prišel iz iskanja po blagovni znamki, neposredno ali prek priporočila: +1
Je zdaj znotraj naravnega okna ponovnega naročila za tisto, kar je kupil: +1

Seštejte jih. Kupci z oceno 3 ali več so vaš segment “z največjo verjetnostjo ponovnega nakupa” — skupina, ki si zasluži zgodnejše, toplejše in bolj natančno sporočilo za drugi nakup. Kupci z oceno 1–2 gredo v počasnejšo negovalno progo, ki zanimanje gradi pred kakršno koli prošnjo. Kupci z oceno 0 ostanejo v lahkem, redkem toku; ne odpisujete jih, le svoje najboljše ponudbe zanje še ne porabljate.

Ilustrativen, a realističen primer. Recimo, da je prejšnji mesec prvi nakup opravilo 1.000 ljudi. Morda jih 220 doseže oceno 3 ali več. To je vaš prednostni segment. Če je vaša izhodiščna pretvorba iz prvega v drugi nakup približno 20 % pri vseh skupaj, potem je usmerjanje najboljšega časa in sporočila na teh 220 — namesto enakomernega razpršenega truda med vseh 1.000 — tam, od koder pridejo dodatna naročila. (Številke so za ponazoritev; vaša dejanska porazdelitev bo drugačna in izračunati bi morali svojo.)

Eno pošteno opozorilo: ocena je izhodiščna stava, ne razsodba. Nekateri z nizko oceno vas bodo presenetili, nekateri z visoko oceno pa ne bodo opravili nakupa. Po nekaj mesecih segment preverite glede na dejanske rezultate in prilagodite uteži. Štirje zgornji signali so smiseln privzetek, a podatki vaše trgovine so avtoriteta — točne vrednosti točk obravnavajte kot nekaj, kar je treba preveriti glede na vaše lastne rezultate, ne kot zakon.

Kaj avtomatizirati okrog segmenta

Segment je uporaben le, če se samodejno nekaj zgodi, ko kupec vanj vstopi ali ga zapusti. Opredelite ga kot živo pravilo, ne kot enkraten izvoz, ki ga nikoli ne ponovite.

Sprožilec: kupčeva ocena preseže 3 (ali izpolni skupne pogoje). Segmenti bi se morali posodabljati, ko se vedenje spreminja — tihi kupec, ki nenadoma odpre tri e-maile in znova obišče stran, bi se moral pomakniti navzgor.
Segment: samo prvi kupci. Ko nekdo opravi drugi nakup, izstopi iz tega segmenta in vstopi v vašo progo za ponovne kupce ali zvestobo. Ne ponujajte drugega nakupa ljudem, ki so ga že opravili.
Časovnica: skupina z visoko verjetnostjo lahko relevantno sporočilo o naslednjem izdelku sliši prej, ker je že angažirana. Počasnejša skupina najprej potrebuje vrednost. Kako to razliko pravilno zadeti, je celotna tema članka kako preprečiti, da drugi nakup zahtevate prezgodaj.
Kanal: e-mail za večino, SMS pa rezerviran za tiste z najvišjimi ocenami, ki so v to privolili — to je vaša najtoplejša publika in vredna dražjega kanala.
Vsebina in cilj: oseben, izdelčno relevanten pozib k smiselnemu naslednjemu naročilu, merjen s stopnjo drugega nakupa znotraj segmenta, ne z odpiranji.

Če je bil vaš prvi izdelek eden od priljubljenih junaških izdelkov, ima odgovor na “kaj naprej” svojo lastno logiko, razloženo v članku kaj ponuditi po nakupu najbolj prodajanega izdelka. In če se odločate, kdaj naj se kupec z visoko oceno premakne v predloge sorodnih izdelkov, to mejo zariše članek kdaj naj prvi kupec vstopi v potek navzkrižne prodaje.

Kako meriti, ali segmentacija deluje

Spremljajte štiri številke in segmente primerjajte med seboj, ne z ničemer:

Stopnja drugega nakupa po segmentu. Skupina z visoko verjetnostjo bi morala v drugo naročilo pretvarjati opazno bolje kot vaša izhodiščna vrednost pri vseh kupcih. Če se to ne zgodi, so vaši točkovalni signali zgrešeni.
Čas do drugega nakupa. Dober segment ne dvigne le pretvorbe, ampak skrajša tudi razmik. Spremljajte, ali prednostna skupina naroči znova hitreje.
Prihodek na prejemnika. Prava preizkušnja, ali se osredotočanje truda izplača. Če najboljše sporočilo pošljete 220 dobro izbranim ljudem, bi moralo prinesti več kot enak trud, razpršen med 1.000.
Stopnja odjav in prijav neželene pošte pri nizki skupini. Če vaša počasnejša proga izgublja stike, prositi preveč in prezgodaj ljudi, ki niso pripravljeni.

Kje se vključi Omnisend

Vse zgoraj je neodvisno od orodja — kupce lahko po potrebi točkujete tudi v preglednici. Sam se pri tem opiram na Omnisend, ker se segmentacija tam izkaže: iz zgodovine naročil, kupljenega izdelka, angažiranosti prek e-maila in SMS-a ter časa od nakupa lahko zgradite živ segment, kupci pa samodejno vstopajo vanj in izstopajo iz njega, ko se njihovo vedenje spreminja. Potem ko sem ga preizkusil proti Klaviyu, je prav ta kombinacija segmentov na podlagi vedenja ter e-maila, SMS-a in potisnih sporočil na enem mestu razlog, da ga uporabljam v svojih trgovinah.

Poštena omejitev: orodje zgradi segment, ne more pa vam povedati, kateri signali so pomembni za vaš katalog. Ta presoja — kateri prvi izdelki, kateri viri, katera okna ponovnega naročila — pride iz analize vaših lastnih ponovnih kupcev. Omnisend je pridruženi partner Shopimationa in ga priporočam iz vsakodnevne uporabe; brezplačni paket zadošča za izgradnjo in preizkus prvega segmenta, preden se odločite zanj.

Vaš naslednji korak

Izpišite prve kupce iz zadnjega četrtletja in vsakega označite s štirimi signali — izdelek, angažiranost, vir, nedavnost. Točkujte jih in poglejte, kako velika je dejansko vaša skupina z oceno 3 in več. Že ta ena razvrstitev vam pove, kam usmeriti najboljši trud za drugi nakup. Od tam določite časovnico s člankom kako preprečiti, da drugi nakup zahtevate prezgodaj, in če želite širši pregled nad tem, katere segmente bi morala vzdrževati vsaka trgovina, članek segmenti kupcev, ki jih mora ustvariti vsaka spletna trgovina tega postavi v kontekst.

First-to-Second Purchase

Segmenting Customers Who Are Most Likely to Buy Again

Posted on July 18, 2026 by

The customers most likely to buy again are rarely the ones who spent the most on their first order. They’re the ones whose first purchase and early behavior match the pattern your existing repeat buyers showed at the same stage. To find them, build a segment from four signals you already have: what they bought, how they paid attention after the order (opens, clicks, site visits), how they were acquired, and how long ago they ordered. Score those signals, group the top slice, and send that group a different, earlier, more personal second-purchase message than everyone else. This page is about how to build that segment — the logic and the fields — not about which product to offer next or when to send it.

Why “everyone who bought once” is the wrong segment

Most stores treat first-time buyers as one bucket. One post-purchase flow, one timeline, one offer, sent to everybody who crossed the line from visitor to customer. It feels fair. It’s also the reason the flow underperforms.

A first-order list is a mix of people with wildly different odds of coming back. Someone who bought a consumable they’ll run out of in five weeks, opened every email since, and came from an organic search for your brand name is a very different prospect from someone who grabbed a one-off gift during a Black Friday promo, from a Meta ad, and hasn’t opened a message since. Send them the same second-purchase push and you either nag the second person into unsubscribing or bore the first person with an offer that arrives too late and asks too little.

The point of segmenting here is narrow: separate the buyers who are already leaning toward a second order from the ones who need more time — or who were never likely to return at all. You treat those groups differently because they are different.

The signals that actually predict a second order

You don’t need a data science team for this. Four signals do most of the work, and every ecommerce platform records them.

1. What they bought first. Product choice is the strongest early tell, because different first products carry different repeat odds. A replenishable item (coffee, supplements, skincare, pet food) has a natural reorder clock built in. An entry-level or “trial size” product signals someone testing you. A deep-discount clearance item often signals a bargain hunter who’ll wait for the next markdown. Sort your catalog by which first products historically lead to a second order — that ranking alone reshapes your segment. Digging into that ranking properly is its own project, covered in how to find which first products create the most loyal customers.

2. Post-purchase engagement. Did they open the shipping email, click the “how to use it” guide, come back to the site, leave a review? Engagement in the two weeks after an order is a live signal of interest. A buyer clicking through your emails is telling you they’re still paying attention. A buyer who’s gone silent isn’t necessarily lost, but they’re not ready for an ask.

3. Acquisition source. Where a customer came from shapes how likely they are to return, sometimes more than what they spent. Branded search and referral traffic tend to produce steadier repeat buyers than cold discount-led ad clicks. If you tag orders with their source, you can weight the segment accordingly — the reasoning behind that gap is worth understanding on its own in why some acquisition channels produce more second purchases than others.

4. Recency and the product’s natural cycle. A customer who bought a 30-day supply 25 days ago is at a very different moment than one who bought the same thing yesterday. Recency only means something against the product’s real reorder window, so read it in context, not as a flat “days since order.”

Turning signals into a workable segment

Here’s the practical build. You don’t need perfect scoring — a rough, honest version beats an elaborate one you never finish.

Start by giving each buyer a simple points tally:

Bought a high-repeat first product: +2
Opened or clicked at least one post-purchase email: +1
Came from branded search, direct, or referral: +1
Is now inside the natural reorder window for what they bought: +1

Add them up. The buyers scoring 3 or higher are your “most likely to buy again” segment — the group worth an earlier, warmer, more specific second-purchase message. Buyers scoring 1–2 go into a slower nurture track that builds interest before any ask. Buyers at 0 stay in a light, low-frequency stream; you’re not writing them off, you’re just not spending your best offer on them yet.

A worked example, illustrative but realistic. Say 1,000 people made a first purchase last month. Maybe 220 of them score 3+. That’s your priority segment. If your baseline first-to-second conversion is around 20% across everyone, concentrating your best timing and message on those 220 — instead of spreading identical effort across all 1,000 — is where the extra orders come from. (Numbers are for illustration; your real distribution will differ, and you should pull your own.)

One honest caveat: a score is a starting bet, not a verdict. Some low scorers will surprise you and some high scorers won’t convert. Review the segment against actual outcomes after a couple of months and adjust the weights. The four signals above are a sensible default, but your store’s data is the authority — treat the exact point values as something to verify against your own results, not a law.

What to automate around the segment

A segment is only useful if something happens automatically when a customer enters or leaves it. Define it as a live rule, not a one-time export you’ll never repeat.

Trigger: a customer’s score crosses 3 (or they meet the combined conditions). Segments should update as behavior changes — a silent buyer who suddenly opens three emails and revisits the site should move up.
Segment: first-time buyers only. Once someone makes a second purchase, they exit this segment and enter your repeat-buyer or loyalty track. Don’t keep pitching a second order to people who already made one.
Timing: the high-likelihood group can hear a relevant next-product message sooner, because they’re already engaged. The slower group needs value first. Getting that gap right is the whole subject of how to avoid asking for the second purchase too early.
Channel: email for most, SMS reserved for the highest scorers who opted in — that’s your warmest audience and worth the pricier channel.
Content and goal: a personal, product-relevant nudge toward a sensible next order, measured by second-purchase rate within the segment, not by opens.

If your first product was one of your popular hero items, the “what next” answer has its own logic, laid out in what to offer after a customer buys a best-selling product. And if you’re deciding when a high scorer should move into related-product suggestions, when should a first-time buyer enter a cross-sell flow draws that line.

How to measure whether the segmentation is working

Watch four numbers, and compare segments against each other rather than against nothing:

Second-purchase rate by segment. The high-likelihood group should convert to a second order noticeably better than your all-buyers baseline. If it doesn’t, your scoring signals are off.
Time to second purchase. A good segment doesn’t only lift conversion, it shortens the gap. Watch whether the priority group reorders faster.
Revenue per recipient. The real test of whether concentrating effort pays. Sending your best message to 220 well-chosen people should out-earn the same effort spread across 1,000.
Unsubscribe and spam rate on the low group. If your slower track is shedding contacts, you’re asking too hard, too soon, for people who aren’t ready.

Where Omnisend fits

Everything above is tool-agnostic — you can score customers in a spreadsheet if you have to. I lean on Omnisend for this because segmentation is where it earns its place: you can build a live segment from order history, product bought, email and SMS engagement, and time since purchase, and have customers flow in and out of it automatically as their behavior changes. After testing it against Klaviyo, that combination of behavior-based segments plus email, SMS, and push in one place is why I run it in my own stores.

The honest limit: the tool builds the segment, but it can’t tell you which signals matter for your catalog. That judgment — which first products, which sources, which reorder windows — comes from looking at your own repeat buyers. Omnisend is an affiliate partner of Shopimation, and I recommend it from daily use; the free tier is enough to build and test a first segment before you commit.

Your next step

Pull last quarter’s first-time buyers and tag each with the four signals — product, engagement, source, recency. Score them, and see how large your 3+ group actually is. That single sort tells you where to point your best second-purchase effort. From there, decide the timing with how to avoid asking for the second purchase too early, and if you want the wider view of which segments every store should maintain, the customer segments every online store should create puts this one in context.