Segmentacija kupcev

Kako segmentirati kupce glede na občutljivost na popuste

Posted on July 19, 2026 by

Segmentiranje kupcev glede na občutljivost na popuste pomeni razdelitev baze v tri skupine na podlagi zgodovine naročil: kupci po polni ceni (malo ali nič njihovih naročil je uporabilo kodo za popust), mešani kupci in kupci, ki jih ženejo ugodnosti (večina ali vsa naročila so imela kodo). Nato promocije usmerjaš temu ustrezno — kupci po polni ceni nehajo prejemati razpošiljanja razprodaj, ki jih niso nikoli potrebovali, popusti pa se skoncentrirajo na ljudi, ki resnično ne bodo kupili brez njih. Poplačilo je marža: vsak kupon, ki ga vnovči nekdo, ki bi plačal polno ceno, je denar, izročen zastonj. Gradnja segmentov vzame eno polje, ki ga večina platform že hrani — ali je naročilo imelo uporabljen popust — plus par poštenih pridržkov o tem, kaj ti ti podatki lahko povejo in česa ne. Tukaj je celotna nastavitev.

Puščanje marže, ki ga tvoje poročilo o kampanji skriva

Razprodajne kampanje v poročilih izgledajo odlično. Prihodek skoči, pretvorbe poskočijo, nadzorna plošča se obarva zeleno. Česar poročilo ne pokaže, je, kdo je vnovčil tiste kode.

Del vsake promocije po celotni strani poberejo kupci, ki so tako ali tako nameravali kupiti, po polni ceni, ta teden. Njihov denar se pokaže v stolpcu prihodka kampanje, zato si promocija pripiše zasluge za naročila, ki jih je zgolj popustila. Če je tvoja marža izdelka 40 % in izvedeš 15 % popusta, se je približno tretjina tvojega dobička na teh naročilih izhlapela — pri kupcih, ki niso potrebovali prepričevanja. Če to počneš mesečno, se sešteva v resnično številko.

Običajen odziv na medle prodaje je globlja ali pogostejša promocija. To obravnava tvoje najboljše kupce in tvoje cenovno najbolj gnane kupce enako, in sčasoma prve vleče proti drugim: ljudje se naučijo, da razprodaja vedno pride, če počakajo. Ta zdrs in kako ga obrniti sta tema kako prodati več, ne da bi kupce navadil čakati na popuste. Segmentacija po občutljivosti je strukturni popravek pod tem.

Najprej pošten pridržek: tvoji podatki so deloma zrcalo

Preden karkoli zgradiš, razumi, kaj zgodovina popustov v resnici meri. Ne meri prirojene cenovne občutljivosti kupca. Meri, kako se je odzval na to, kar si mu poslal.

Štejeta dve popačitvi:

Morda si sam ustvaril svoje »z ugodnostmi gnane« kupce. Če je kupčevo vsako naročilo uporabilo kodo, a je vsaka e-pošta, ki si mu jo kdaj poslal, vsebovala kodo, mu ni bila niti enkrat ponujena priložnost, da bi plačal polno ceno. Njegovo vedenje je tvoje vedenje, odsevano nazaj.
Razprodaje po celotni strani onesnažijo signal. Naročilo, oddano med tvojim dogodkom črnega petka, ne pove skoraj ničesar — vsi so dobili tisto ceno. Naročila, kjer je kupec šel iskat kodo ali kupil šele po e-pošti s kuponom, so veliko bolj zgovorna.

Nobena od teh težav pristopa ne ubije. Le pomeni, da segment obravnavaš kot delovno hipotezo in pustiš prihodnjim pošiljanjem po polni ceni, da jo potrdijo ali popravijo. Kupec, ki kupi iz kampanje brez popusta, se povzpne ven iz skupine, ki jo ženejo ugodnosti, segment pa postaja resničnejši vsak mesec.

Kako zgraditi tri segmente

Gradivo za pravilo je delež kupčevih naročil, ki so uporabila popust. Večina e-poštnih platform zna filtrirati po »naročilo vsebuje popust« ali po skupni vrednosti popusta na naročilo; če tvoja ne zna, te izvoz v preglednico s stolpcem popusta pripelje do cilja.

Kupci po polni ceni: dve ali več naročil, s kodami za popust na 0–25 % od njih. Ti kupujejo na podlagi izdelka, časa in zaupanja.
Mešani kupci: kode na približno 25–75 % naročil. Plačali bodo polno ceno za pravo stvar in vzeli ugodnost, ko je ponujena.
Kupci, ki jih ženejo ugodnosti: kode na 75 %+ naročil. Brez kode, brez naročila — vsaj doslej.

Dve opombi h gradnji. Zahtevaj vsaj dve naročili, preden kogar koli razvrstiš; eno samo prvo naročilo s pozdravno kodo ne dokaže ničesar, razen da tvoj pozdravni tok deluje. In pri štetju izloči naročila iz svojih največjih dogodkov po celotni strani, če tvoja platforma to dopušča, iz zgoraj navedenega razloga. Enkratni kupci in potencialne stranke, ki niso nikoli kupile, stojijo zunaj te sheme — bolje jih je obravnavati skozi segmentacijo po nakupnem vedenju na splošno.

Odstotki so stvar presoje, ne zakoni. Trgovina, ki popuščanja redko, jih lahko zaostri; trgovina, ki okreva po letih tedenskih promocij, naj pričakuje, da bo vedro z ugodnostmi gnanih sprva sramotno veliko.

Kaj naj prejme vsak segment

Tu se marža prihrani ali potroši.

Kupci po polni ceni nehajo dobivati razprodajne kampanje. Pika. Dobijo nove prihode, ponovne zaloge, zgodnji dostop, vsebino, ugodnosti za zvestobo — vse razen odstotka popusta. Če se zdi tvegano preskočiti jih pri veliki promociji, jim daj isti dogodek, uokvirjen drugače: zgodnji dostop po polni ceni ali darilo ob nakupu, ki te stane nabavne cene blaga namesto čiste marže. Ti kupci se močno prekrivajo s tvojimi najboljšimi kupci in si zaslužijo obravnavo, opisano v kako zgraditi segment kupcev z visoko vrednostjo.

Mešani kupci dobijo tvoj običajen koledar: večinoma pošiljanja po polni ceni, plus pristne sezonske promocije. Opazuj njihovo premikanje. Če tvoja mešana skupina kar naprej seli proti tistim, ki jih ženejo ugodnosti, je tvoj koledar preveč promocijski — to je diagnoza, ne smola.

Kupci, ki jih ženejo ugodnosti, so tam, kamor spadajo popusti, in kjer si lahko kirurško natančen. Sprožilec: tvoja načrtovana promocija. Segment: samo tisti, ki jih ženejo ugodnosti. Kanal: e-pošta, s SMS-om za sklepne ure, če imaš privolitev. Vsebina: ponudba, navedena naravnost — to občinstvo se odziva na jasnost, ne na pripovedovanje zgodb. Cilj: prihodek, ki ga sicer sploh ne bi dobil. Ker gre ponudba samo ljudem, ki ne pretvorijo brez nje, jo lahko celo izvedeš malce globljo, kot bi bila razprodaja po celotni strani, in vseeno skupno podariš manj marže.

Ena meja, ki jo je vredno postaviti: drži tudi samodejne popuste na kratkem povodcu. Kupon za opuščeno košarico, ki se sproži za vse, tiho vgrajuje občutljivost v tvoj celoten seznam — kdaj naj dodaš popust v tok za opuščeno košarico obravnava, kako ga zamejiti.

Preprosta primerjava marže (ponazoritvena)

Recimo 2.000 kupcev, povprečno naročilo 60 €, marža izdelka 40 % (24 €). Izvedeš 15 % popusta (9 € na naročilo).

Razpošiljanje po celotni strani: 200 naročil — 80 od ljudi, ki bi kmalu tako ali tako kupili po polni ceni. Strošek teh 80 nepotrebnih popustov: 720 € čiste marže, več kot tretjina celotne podaritve promocije.

Segmentirana različica: ponudba doseže le ~600 kupcev, ki jih ženejo ugodnosti, in izvleče morda 90 naročil, ki sicer ne bi obstajala. Preostalih 110 razprodajnih naročil ne izgine — mnogi od teh kupcev še vedno kupijo iz kampanj po polni ceni, ki jih namesto tega prejmejo, pri 24 € marže namesto 15 €. Manj vnovčenih kuponov, podobno skupno število naročil, vidno boljša mešana marža. (Vsaka številka tukaj je ponazoritvena — razmerja so odvisna povsem od tvoje trgovine, kar je natanko razlog, da zgradiš segmente in preveriš.)

Kako to meriti

Preskoči nečimrne odpre. Zgodbo povedo štiri številke:

Mešana marža na cikel kampanje — skupna vrednost podarjenega popusta na mesec, deljena s prihodkom. Ta naj pada.
Delež naročil po polni ceni med segmentoma po polni ceni in mešanim skozi čas. Naraščanje je zmaga.
Prihodek na prejemnika za pošiljanja tistim, ki jih ženejo ugodnosti, v primerjavi s tvojimi starimi razpošiljanji po celotni strani.
Selitev segmentov — krčenje vedra z ugodnostmi gnanih, ko poboljšani kupci kupujejo brez kod, je signal uspeha na dolgi rok.

Če želiš okvir za presojanje celotnega programa namesto ene kampanje, gre kako izmeriti, ali segmentacija izboljša prihodek globlje.

Kje se vključi Omnisend

Te segmente lahko zgradiš v katerikoli platformi, ki hrani podatke o popustih na naročilo. V svojih trgovinah uporabljam Omnisend — izbran namesto Klaviya po tem, ko sem preizkusil oba, večinoma zaradi cene in tega, kako malo časa vzame vzdrževanje segmentov — in deli, ki tukaj štejejo, so filtri segmentov na poljih popusta naročila, samodejne posodobitve segmentov ob prihodu novih naročil in občinstva kampanj, ki jih lahko nastaviš tako, da vključujejo z ugodnostmi gnane kontakte, medtem ko v istem pošiljanju izključujejo tiste po polni ceni.

Poštene meje: tvoja platforma za spletno trgovino mora podatke o popustih posredovati čisto, zgodovinska razvrstitev je le tako dobra kot sinhronizirana zgodovina naročil, in nobeno orodje ne popravi promocijskega koledarja, ki popušča vsak teden. Segment ti pove, kdo je občutljiv; odločitev, da boš poslal manj kuponov, je še vedno živec, ki ga moraš obdržati.

Tvoj naslednji korak

Izvleci naročila zadnjega četrtletja in odgovori na eno vprašanje: kolikšen odstotek tvojega prihodka je nosil kodo za popust in koliko od tega je prišlo od ponovnih kupcev, ki so kupovali tudi po polni ceni? Ta ena sama številka običajno razjasni, ali je to stranski projekt ali izredne razmere. Nato najprej zgradi segment po polni ceni — zaščititi maržo, ki jo že zaslužiš, prekaša optimiziranje popustov, ki jih podarjaš.

Customer Segmentation

Segmenting Customers by Discount Sensitivity

Posted on July 17, 2026July 17, 2026 by

Segmenting customers by discount sensitivity means splitting your base into three groups based on order history: full-price buyers (few or none of their orders used a discount code), mixed buyers, and deal-driven buyers (most or all orders came with a code). Then you route your promotions accordingly — full-price buyers stop receiving sale blasts they never needed, and discounts get concentrated on the people who genuinely won’t buy without one. The payoff is margin: every coupon redeemed by someone who would have paid full price is money handed back for nothing. Building the segments takes one field most platforms already store — whether an order had a discount applied — plus a couple of honest caveats about what that data can and can’t tell you. Here’s the whole setup.

The margin leak your campaign report hides

Sale campaigns look great in reports. Revenue spikes, conversion jumps, the dashboard turns green. What the report doesn’t show is who redeemed those codes.

A chunk of every sitewide promotion gets claimed by customers who were going to buy anyway, at full price, that week. Their money shows up in the campaign’s revenue column, so the promotion takes credit for orders it merely discounted. If your product margin is 40% and you run 15% off, roughly a third of your profit on those orders evaporated — on buyers who needed no persuading. Do that monthly and it compounds into a real number.

The usual response to soft sales is a deeper or more frequent promotion. That treats your best customers and your most price-driven customers identically, and over time it drags the first group toward the second: people learn that a sale always comes if they wait. That slide, and how to reverse it, is the subject of how to sell more without training customers to wait for discounts. Sensitivity segmentation is the structural fix underneath it.

First, an honest caveat: your data is partly a mirror

Before you build anything, understand what discount history actually measures. It doesn’t measure a customer’s inborn price sensitivity. It measures how they responded to what you sent them.

Two distortions matter:

You may have created your “deal-driven” customers. If a customer’s every order used a code, but every email you ever sent them contained a code, they’ve never once been offered the chance to pay full price. Their behavior is your behavior, reflected back.
Sitewide sales pollute the signal. An order placed during your Black Friday event says almost nothing — everyone got that price. Orders where a customer went hunting for a code, or bought only after a coupon email, are far more telling.

Neither problem kills the approach. It just means you treat the segment as a working hypothesis and let future full-price sends confirm or correct it. A customer who buys from a no-discount campaign gets promoted out of the deal-driven group, and the segment gets truer every month.

How to build the three segments

The rule material is the share of a customer’s orders that used a discount. Most email platforms can filter on “order contains discount” or on total discount value per order; if yours can’t, a spreadsheet export with the discount column gets you there.

Full-price buyers: two or more orders, with discount codes on 0–25% of them. These people buy on product, timing, and trust.
Mixed buyers: codes on roughly 25–75% of orders. They’ll pay full price for the right thing and take a deal when offered.
Deal-driven buyers: codes on 75%+ of orders. No code, no order — at least so far.

Two build notes. Require at least two orders before classifying anyone; a single first order with a welcome code proves nothing except that your welcome flow works. And exclude orders from your biggest sitewide events when counting, if your platform allows it, for the reason above. One-time buyers and prospects who’ve never purchased sit outside this scheme — they’re better handled through purchase-behavior segmentation generally.

The percentages are judgment calls, not laws. A store that discounts rarely can tighten them; a store recovering from years of weekly promos should expect the deal-driven bucket to be embarrassingly large at first.

What each segment should receive

This is where the margin gets saved or spent.

Full-price buyers stop getting sale campaigns. Full stop. They get new arrivals, restocks, early access, content, loyalty perks — everything except a percentage off. If skipping them on a big promotion feels risky, give them the same event framed differently: early access at full price, or a gift with purchase, which costs you cost-of-goods instead of straight margin. These customers overlap heavily with your best customers, and they deserve the treatment described in how to build a high-value customer segment.

Mixed buyers get your normal calendar: mostly full-price sends, plus the genuine seasonal promotions. Watch their drift. If your mixed group keeps migrating toward deal-driven, your calendar is too promotional — that’s a diagnosis, not bad luck.

Deal-driven buyers are where discounts belong, and where you can be surgical. Trigger: your planned promotion. Segment: deal-driven only. Channel: email, with SMS for the closing hours if you have consent. Content: the offer, stated plainly — this audience responds to clarity, not storytelling. Goal: revenue you would otherwise not get at all. Because the offer goes only to people who don’t convert without one, you can even run it slightly deeper than a sitewide sale would have been, and still give away less margin in total.

One boundary worth setting: keep automated discounts on the same leash. An abandoned-cart coupon that fires for everyone quietly trains sensitivity into your whole list — when should you add a discount to an abandoned cart flow covers how to gate it.

A simple margin comparison (illustrative)

Say 2,000 customers, average order €60, product margin 40% (€24). You run 15% off (€9 per order).

Sitewide blast: 200 orders — 80 from people who’d have bought full price soon anyway. Cost of those 80 needless discounts: €720 of pure margin, over a third of the promotion’s total giveaway.

Segmented version: the offer reaches only the ~600 deal-driven customers, pulling maybe 90 orders that wouldn’t otherwise exist. The other 110 sale-orders don’t vanish — many of those buyers still purchase from the full-price campaigns they receive instead, at €24 margin rather than €15. Fewer coupons redeemed, similar total orders, visibly better blended margin. (Every number here is illustrative — the ratios depend entirely on your store, which is exactly why you build the segments and check.)

How to measure it

Skip vanity opens. Four numbers tell the story:

Blended margin per campaign cycle — total discount value given away per month, divided by revenue. This should fall.
Full-price order share among the full-price and mixed segments over time. Rising is winning.
Revenue per recipient for deal-driven sends versus your old sitewide blasts.
Segment migration — the deal-driven bucket shrinking as reformed customers buy without codes is the long-game success signal.

If you want a framework for judging the whole program rather than one campaign, how to measure whether segmentation improves revenue goes deeper.

Where Omnisend fits

You can build these segments in any platform that stores discount data per order. I use Omnisend in my own stores — chosen over Klaviyo after trying both, mostly for price and how little time segment maintenance takes — and the pieces that matter here are segment filters on order discount fields, automatic segment updates as new orders arrive, and campaign audiences you can set to include deal-driven contacts while excluding full-price ones in the same send.

The honest limits: your ecommerce platform has to be passing discount data through cleanly, historical classification is only as good as the order history synced, and no tool fixes a promotional calendar that discounts every week. The segment tells you who’s sensitive; deciding to send fewer coupons is still a nerve you have to hold.

Your next step

Pull last quarter’s orders and answer one question: what percentage of your revenue carried a discount code, and how much of it came from repeat customers who’ve also bought at full price? That single number usually settles whether this is a side project or an emergency. Then build the full-price segment first — protecting the margin you already earn beats optimizing the discounts you give away.