Spletna potisna obvestila

Kako uporabiti potisna obvestila ob znižanju cene

Posted on July 18, 2026 by

Obvestilo ob znižanju cene določenemu naročniku pove, da je določen izdelek, za katerega je pokazal zanimanje — si ga ogledal, ga dal v košarico, ga shranil na seznam želja — zdaj cenejši. Poslano na ta način je eno najbolj dobrodošlih obvestil, ki jih trgovina lahko pošlje: oseba je artikel že želela, spremenil se je le pomislek, ki se je skrčil. Pravila, ki ga ohranjajo delujočega: opozorite le zainteresirani segment, nikoli celotnega seznama; sprožite le ob znižanjih, dovolj velikih, da štejejo; izključite vsakogar, ki je artikel nedavno kupil; in opozorite le na znižanja cen, ki ste jih tako ali tako izvajali. To je opozorilo ena na ena, ki ga sproži sprememba cene želenega izdelka. Trgovinsko promocijo, razposlano vsem, je druga igra z drugačno matematiko — to je potisna obvestila za bliskovite razprodaje — »spet je na voljo« pa je obvestilo o ponovni zalogi, ki odgovarja na razpoložljivost, ne na ceno.

Problem, ki ga to rešuje: zainteresirani ljudje ne preverjajo znova

Vsaka trgovina nosi plast skoraj-strank: ljudi, ki so našli izdelek, jim je bil všeč in so obtičali pri ceni. Nekateri so zapustili košarico pri znesku. Nekateri so si stran ogledali trikrat in je nikoli niso dali v košarico. Niso rekli ne. Rekli so ne po tej ceni.

Nato, tedne kasneje, to ceno znižate — sezonsko znižanje, razprodaja, ugoden dobaviteljski posel — in skoraj nihče od teh ljudi za to ne izve. Ne osvežujejo vaših strani izdelkov in čakajo; tega ne počne nihče. Znižanje se prikaže komurkoli, ki tisti teden slučajno obišče, kar je večinoma nov promet, ki sprva sploh ni imel pomisleka glede cene. Prav skupina, za pretvorbo katere je bil popust ustvarjen, za to nikoli ne sliši.

To je puščanje. Maržo že trošite. Le pred napačnim občinstvom jo trošite.

Zakaj kampanjska e-pošta o znižanju tega ne pokrije

Običajna poteza je kampanja »RAZPRODAJA« celotnemu seznamu, in ta ima svoje mesto. A za to nalogo je napačno orodje, iz treh razlogov. Doseže vaš seznam, ne vaših ljudi, zainteresiranih za ta izdelek — večinoma različni skupini. Prispe po urniku kampanje, ne ob trenutku spremembe cene, znižano zaloga v priljubljeni velikosti pa ne čaka na četrtkove novice. In oseba, ki nikoli ni pokazala zanimanja za znižani artikel, doživlja vašo razprodajno e-pošto kot šum, kar je način, kako se utrujenost od seznama kopiči.

Obvestilo ob znižanju cene vse tri obrne: pravi človek, ob trenutku spremembe in le ljudem, za katere je to res dobra novica. Prav zadnji del je razlog, zakaj so ta opozorila običajno najbolje sprejeta avtomatizirana obvestila, ki jih trgovina pošlje: naročnik je, v resnici, že prosil, da to sliši.

Matematika smiselnega znižanja

Nastavite prag. Sprememba za 2 € pri izdelku za 40 € ni obvestilo, opozarjanje na nepomembna nihanja pa vaš proračun za prekinitve zapravi za nič. Nekje okoli 10–15 % popusta od cene, ki jo je naročnik prvotno videl, je razumno dno (ilustrativno — svojega nastavite po marži in običajni globini popustov). Pri nizkih cenah odstotek šteje bolj kot absolutna vrednost; pri visokih cenah pa obrnite — znižanje za 60 € pri izdelku za 600 € je vredno obvestila, tudi če je to 10 %.

En tih izračun, ki ga je vredno narediti pred nastavitvijo praga: vaša ničelna točka pri popustu. Če je vaša marža 40 % in ceno znižate za 15 %, potrebujete občutno več pretvorb iz opozorjenega segmenta, kot bi jih dobili pri polni ceni, da se vaja izplača. Ker obvestila ob znižanju cene gredo le ljudem, ki so že obtičali pri polni ceni, dodatna pretvorba običajno preseže to letvico — a »običajno« v tem stavku opravlja delo, kar je razlog, zakaj je razdelek o merjenju spodaj pomemben.

Zgradite tok

Sprožilec: znižanje cene čez vaš prag pri izdelku, za katerega je naročnik pokazal zanimanje.

Segment: naročniki, ki so si izdelek ogledali (nedavno — postavite okno, recimo zadnjih 30–60 dni), ga dali v košarico brez nakupa ali ga shranili na seznam želja. Signala seznama želja in košarice sta najmočnejša; sprožilce na podlagi ogleda kvalificirajte enako skrbno, kot bi jih pri zapuščenem ogledu, da naključen klik nekoga ne uvrsti v opozorilni ribnik.

Izključitve — tiste, ki jih ljudje pozabijo:
– Vsak, ki je artikel kupil. To je najbolj kislo obvestilo, ki ga trgovina lahko pošlje: povedati kupcu prejšnjega tedna, da je njegov nakup zdaj 20 % cenejši. V najboljšem primeru so razdraženi. Na trgih s pričakovanji glede prilagoditve cene pa so v vašem predalu za podporo in prosijo za vračilo razlike. Izključite kupce izdelka, pika, in izključitvi dajte velikodušen pogled nazaj.
– Vsak, ki ima artikel prav zdaj v aktivni košarici — ti sodijo v vaš tok za zapuščeno košarico, obe avtomatizaciji pa se ne bi smeli nakopičiti na isto osebo v istem popoldnevu.

Časovni razpored: v nekaj urah po spremembi cene. Hitrost je celotna premisa; če opozorilo lahko prispe dan prepozno, razprodajne zaloge v priljubljenih velikostih morda ne.

Sporočilo: začnite z izdelkom in novo številko. »Znižanje cene: [ime izdelka] je zdaj 89 € (bilo 119 €).« Konkretno premaga zvijačno — »nekaj, kar ste želeli, je pravkar postalo cenejše« sili ljudi v ugibanje, ugibanje pa izgublja klike. Z globoko povezavo peljite na stran izdelka. Eno obvestilo na znižanje na izdelek; brez naknadnih opomnikov »še vedno znižano!«.

Past: učenje ljudi, naj čakajo

Bodite iskreni glede spodbude, ki jo ustvarjate. Če vsak izdelek, pri katerem se obiskovalec zadrži, sčasoma sproži popust, se bodo vaše najbolj prebrisane stranke naučile igre: brskaj, shrani na seznam želja, čakaj. Zgradili boste stroj, ki polnocenovno namero pretvarja v znižana naročila.

Tri varovalke preprečijo, da bi se tok skisal: opozarjajte le na znižanja, ki ste jih tako ali tako izvajali iz trgovinskih razlogov — nikoli ne znižajte cene, ker se je na njej nabralo opazovalcev. Naj bo okno zanimanja tesno, da se vzorec »čakaj dovolj dolgo in bo cenejše« redko dokonča. In omejite, koliko obvestil o znižanju cene kateri koli naročnik dobi v mesecu; nekdo, ki jih prejme štiri, je učen, ne glede na vaše namene.

Primer iz trgovine

Ilustrativna skica: trgovina z opremo za zimske športe konča sezono s smučarsko jakno za 260 €, ki se počasi prodaja, in jo zniža na 199 €. Namesto da bi to zakopala v marčevsko razprodajno e-pošto, sprememba cene sproži obvestilo 120 naročnikom, ki so si to jakno ogledali ali jo dali v košarico od decembra, minus 14, ki so jo kupili. »Znižanje cene: [ime jakne] je zdaj 199 € (bilo 260 €). Ostale velikosti M–XL.« Tudi ob skromni stopnji pretvorbe iz tega toplega, drobnega segmenta se razprodajna zaloga v urah spremeni v naročila, prodana točno tistim ljudem, katerih pomislek je bila stara cena — brez porabe pozornosti kogar koli drugega.

Kako meriti uspešnost

Stopnja klikov na obvestila ob znižanju cene — ta bi morala biti pri vrhu vseh vaših avtomatizacij; če ni, je vaš segment zanimanja preohlapen.
Stopnja pretvorbe opozorjenih naročnikov na znižanem izdelku, najbolje v primerjavi s tem, kako se to znižanje odreže z rednim prometom. Vrzel je pravi prispevek toka.
Prihodek in marža na opozorilo — spremljajte maržo poleg prihodka, saj vsako od teh naročil po definiciji nosi popust.
Vedenje ob ponovnih opozorilih — pazite na naročnike, ki večkrat kupijo šele po opozorilih. Nekaj jih je v redu. Rastoča skupina pomeni, da je treba zaostriti prage ali okna.

Pripisovanje postane megleno, ko se znižanje pojavi tudi v e-pošti in na pasicah na spletnem mestu, zato se vnaprej odločite, kako boste pripisovali potisnim obvestilom — praktičen pristop je v članku kako meriti prihodek iz spletnih potisnih obvestil.

Nastavitev v Omnisendu

To poganjam v Omnisendu, ker znižanje cene stoji poleg sprožilcev za košarico, ogled in ponovno zalogo v istem graditelju, in izključitve so tam, kjer to šteje: tok lahko pred pošiljanjem preveri zgodovino nakupov in trenutno vsebino košarice, kar je natanko logika, ki prepreči obvestilo »kupili ste ga, zdaj je cenejši«. Iskrenost glede nastavitve: tok je odvisen od tega, da vaš vir izdelkov spremembe cen sinhronizira pravočasno, zato po izgradnji spremenite ceno testnega izdelka in sami potrdite časovni razpored, preden ga zaupate resničnim naročnikom.

Vaš naslednji korak

Poiščite svoje zadnje resnično znižanje in odgovorite na eno vprašanje: koliko ljudi, ki so si ta izdelek ogledali ali ga dali v košarico, je kdaj izvedelo, da se je cena spremenila? Če je odgovor »kdorkoli je slučajno obiskal«, zgradite tok zgoraj pred svojim naslednjim ciklom znižanj — popust vas maržo že stane; opozorilo je tisto, zaradi katerega si prisluži pretvorbe. Nato uporabite isto logiko opozarjanja na želen izdelek za razpoložljivost z obvestili o ponovni zalogi, ki isto toplo občinstvo pretvorijo, ne da bi porabili niti cent marže.

Web Push Notifications

How to Use Push Notifications for Price Drops

Posted on July 17, 2026 by

A price drop push tells a specific subscriber that a specific product they showed interest in — viewed it, carted it, wishlisted it — is now cheaper. Sent that way, it’s one of the most welcome notifications a store can send: the person already wanted the item, and the only thing that changed is the objection shrinking. The rules that keep it working: alert only the interested segment, never the whole list; only fire on drops big enough to matter; exclude anyone who recently bought the item; and only alert price cuts you were making anyway. This is a one-to-one alert triggered by a price change on a wanted product. A store-wide promotion blasted to everyone is a different play with different math — that’s push for flash sales — and “it’s available again” is a back-in-stock alert, which answers availability, not price.

The problem this solves: interested people don’t check back

Every store carries a layer of almost-customers: people who found the product, liked it, and stalled on price. Some abandoned a cart at the total. Some viewed the page three times and never carted. They didn’t say no. They said not at this price.

Then, weeks later, you cut that price — a seasonal markdown, a clearance, a supplier deal — and almost none of those people find out. They aren’t refreshing your product pages waiting; nobody does that. The markdown gets shown to whoever happens to visit that week, which is mostly new traffic that never had the price objection in the first place. The exact group the discount was built to convert never hears about it.

That’s the leak. You’re already spending the margin. You’re just spending it in front of the wrong audience.

Why a markdown campaign email doesn’t cover it

The usual move is a “SALE” campaign to the whole list, and it has its place. But it’s the wrong tool for this job, for three reasons. It reaches your list, not your interested-in-this-product people — mostly different groups. It arrives on the campaign’s schedule, not at the moment of the price change, and marked-down inventory in a popular size doesn’t wait for Thursday’s newsletter. And a person who never showed interest in the discounted item experiences your sale email as noise, which is how list fatigue compounds.

The price drop push inverts all three: right person, at the moment of the change, and only to people for whom this is genuinely good news. That last part is why these alerts tend to be the best-received automated pushes a store sends: the subscriber had, in effect, already asked to hear this.

The math of a meaningful drop

Set a threshold. A €2 change on a €40 product isn’t a notification, and alerting on trivial fluctuations spends your interruption budget on nothing. Somewhere around 10–15% off the price the subscriber originally saw is a reasonable floor (illustrative — set yours by your margins and typical discount depth). Percentage matters more than absolute value at low prices; at high prices, flip it — a €60 cut on a €600 item is worth a push even though it’s 10%.

One quiet calculation worth doing before you set the threshold: your break-even on the discount. If your margin is 40% and you drop the price 15%, you need meaningfully more conversions from the alerted segment than you’d have gotten at full price for the exercise to pay. Since price drop pushes only go to people who already stalled at full price, the incremental conversion usually clears the bar — but “usually” is doing work in that sentence, which is why the measurement section below matters.

Build the flow

Trigger: price decrease beyond your threshold on a product a subscriber has shown interest in.

Segment: subscribers who viewed the product (recently — put a window on it, say the last 30–60 days), carted it without buying, or wishlisted it. Wishlist and cart signals are strongest; qualify view-based triggers the same careful way you would for browse abandonment, so a stray click doesn’t put someone in the alert pool.

Exclusions — the ones people forget:
– Anyone who bought the item. This is the single sourest push a store can send: telling last week’s buyer their purchase is now 20% cheaper. Best case they’re irritated. In markets with price-adjustment expectations, they’re in your support inbox asking for a refund of the difference. Exclude buyers of the product, full stop, and give the exclusion a generous lookback.
– Anyone with the item in an active cart right now — they belong to your abandoned cart flow, and the two automations shouldn’t stack on the same person in the same afternoon.

Timing: within a few hours of the price change. Speed is the entire premise; if the alert can arrive a day late, the sale inventory in popular sizes may not.

Message: lead with the product and the new number. “Price drop: the

is now €89 (was €119).” Concrete beats coy — “something you wanted just got cheaper” makes people guess, and guessing loses clicks. Deep-link to the product page. One push per drop per product; no follow-up “still discounted!” reminders.

The trap: teaching people to wait

Be honest about the incentive you’re creating. If every product a visitor lingers on eventually pushes them a discount, your sharpest customers will learn the game: browse, wishlist, wait. You’ll have built a machine that converts full-price intent into discounted orders.

Three guardrails keep the flow from souring: only alert markdowns you were making anyway, for merchandising reasons — never drop a price because watchers accumulated on it. Keep the interest window tight, so the “wait long enough and it’s cheaper” pattern rarely completes. And cap how many price drop pushes any subscriber gets in a month; someone receiving four is being trained, whatever your intentions.

A store example

Illustrative sketch: a store selling winter sports gear ends the season with a €260 ski jacket moving slowly and marks it down to €199. Instead of burying that in a March clearance email, the price change fires a push to the 120 subscribers who viewed or carted that jacket since December, minus the 14 who bought it. “Price drop: the [jacket name] is now €199 (was €260). Sizes M–XL left.” Even at a modest conversion rate off that warm, tiny segment, that’s clearance inventory turning into orders within hours, sold to the exact people whose objection was the old price — with nobody else’s attention spent.

How to measure it

Click rate on price drop pushes — this should be near the top of all your automations; if it isn’t, your interest segment is too loose.
Conversion rate of alerted subscribers on the discounted product, ideally against how that markdown performs with regular traffic. The gap is the flow’s real contribution.
Revenue and margin per alert — track margin alongside revenue, since every one of these orders carries a discount by definition.
Repeat-alert behavior — watch for subscribers who repeatedly buy only after alerts. A few is fine. A growing cohort means your thresholds or windows need tightening.

Attribution gets murky when a markdown also appears in emails and on-site banners, so decide upfront how you’ll credit push — the practical approach is in how to measure revenue from web push notifications.

Setting it up in Omnisend

I run this in Omnisend because price drop sits alongside cart, browse, and back-in-stock triggers in the same builder, and the exclusions are where that matters: the flow can check purchase history and current cart contents before sending, which is exactly the logic that prevents the “you bought it, now it’s cheaper” push. Setup honesty: the flow depends on your product feed syncing price changes promptly, so after building it, change a test product’s price and confirm the timing yourself before trusting it with real subscribers.

Your next step

Find your last real markdown and answer one question: how many people who’d viewed or carted that product ever learned the price changed? If the answer is “whoever happened to visit,” build the flow above before your next markdown cycle — the discount is already costing you margin; the alert is what makes it earn conversions. Then apply the same wanted-product-alert logic to availability with back-in-stock push notifications, which converts the same warm audience without spending a cent of margin.