Naročniški in potrošni izdelki

Zakaj fiksni opomniki za ponovni nakup ne delujejo, ko se poraba razlikuje

Posted on July 19, 2026 by

Fiksni opomnik za ponovni nakup — “pošlji e-pošto vsem 30 dni po nakupu” — ne deluje, ker tvoji kupci izdelka ne porabljajo z isto hitrostjo, tako da je en sam interval pravilen za skoraj nobenega od njih. Za redke uporabnike je prezgoden, za intenzivne prepozen. Prezgodnji opomniki se ignorirajo in izučijo ljudi, da te preslišijo; prepozni pridejo, ko je kupcu že zmanjkalo in je kupil drugje. Ena številka se lahko ujema le s povprečnim kupcem, v večini kategorij potrošnih izdelkov pa je razpon okoli tega povprečja dovolj širok, da “povprečje” opisuje le majhno manjšino. Ta zapis pojasni, zakaj se fiksni model podre, koliko prihodka tiho stane in kaj storiti namesto tega.

Situacija, ki jo verjetno prepoznaš

Opomnik za ponovno naročilo si nastavil, ker se je zdelo očitno: ljudje porabijo izdelek, torej jih opomni, naj kupijo znova. Izbral si zamik — 30 dni se zdi razumno — in ga vklopil. Nekaj naredi. Nekaj ponovnih naročil pricurlja. A stopnja ponovnih nakupov je nižja, kot si upal, odjave pri tem pošiljanju se plazijo navzgor in ne moreš se otresti občutka, da veliko kupcev ostaja brez in se ne vrne nikoli več.

Res je. Opomnik ni pokvarjen; pokvarjena je predpostavka pod njim. Fiksni zamik predpostavlja, da vsi porabljajo z eno hitrostjo, potrošni izdelki pa so točno kategorija, kjer ta predpostavka razpade.

Zakaj se poraba razlikuje bolj, kot bi si mislil

Pretehtaj razloge, zakaj dve osebi porabita isti izdelek z različno hitrostjo:

Velikost gospodinjstva. Enemu človeku šampon zdrži gospodinjstvo štirih štirinajst dni, samski osebi pa dva meseca.
Odmerek in navada. Dva kupca dopolnil, ista steklenička — eden jemlje odmerek z embalaže, eden ga podvoji, eden polovico časa pozabi.
Velikost pakiranja. Kupec, ki je kupil vrednostni večpak, je na povsem drugi uri kot tisti, ki je kupil enkratnega, čeprav sta “isti” kupec, ki kupuje “isti” izdelek.
Sezonskost. Krema za sončenje julija in krema za sončenje novembra pri isti osebi nista porabljeni z isto hitrostjo. Prav tako ne sprej proti alergijam ali mešanica elektrolitov. Ta vzorec si zasluži svojo obravnavo — glej obvladovanje sezonskih sprememb v porabi izdelkov.

Sestavi to skupaj in razpon resničnih intervalov ponovnega naročanja za en sam izdelek zlahka seže od dveh tednov do dveh mesecev. 30-dnevni opomnik sedi na sredini tega razpona in je res dobro odmerjen le za tanko rezino ljudi blizu povprečja.

Kje denar dejansko pušča

Tukaj je del, ki ga je zlahka spregledati: fiksni opomnik odpove v dveh smereh hkrati in obe te staneta.

Prepozen za intenzivne uporabnike — izgubljene prodaje. Kupci, ki izdelek porabijo najhitreje, so po definiciji tvoji najboljši kupci potrošnih izdelkov. Ponovno bi naročali najpogosteje in v enem letu porabili največ. Opomnik, ki pride, potem ko so že ostali brez, jih doseže, ko so si že našli nadomestek. Svoje najvrednejše kupce izgubljaš zaradi slabe časovnice, kar je najhujša možna skupina za izgubiti.

Prezgoden za redke uporabnike — utrujenost seznama. Opomnik, ki pade, ko ima nekdo še dve tretjini, se bere kot nadlegovanje. Ignorira ga, potem ignorira naslednjega, potem se odjavi ali te označi kot neželeno pošto. Zdaj tvoja dostavljivost pade za vse, ti pa si poškodoval kanal, da bi rešil opomnik, ki tega človeka tako ali tako ni imel namena pretvoriti.

Groba ilustracija: predstavljaj si 1.000 kupcev izdelka, katerih resnični intervali ponovnega naročanja se raztezajo od 15 do 60 dni. Eno pošiljanje na 30. dan bi lahko padlo “v pravo okno” morda za kakih nekaj sto od njih. Preostali ga dobijo prezgodaj ali prepozno. Že skromno izboljšanje časovnice pri preostalih 800 — pretvorba dela zamujenih ponovnih naročil v dejanska ponovna naročila — običajno zasenči vse, kar bi dobil z lepšo predlogo ali popustom. Številke so ilustrativne, a asimetrija je resnična: v tem toku časovnica premaga kreativo.

Zakaj običajne “rešitve” tega ne rešijo

Pošlji ga preprosto prej. Opomnik premakni na 20. dan in rešiš nekaj intenzivnih uporabnikov, a nadleguješ še več redkih. Odpoved si premaknil, ne odstranil.

Pošlji več opomnikov. Opomnik na 20., 30. in 45. dan pokrije več razpona, a zdaj redki uporabniki dobijo tri e-pošte, ki jih ne potrebujejo, in problem utrujenosti si poslabšal. Količina ni nadomestek za natančnost. Če boš že poslal kratko serijo, mora biti premišljeno odmerjena — kako zgraditi pot ponovnega naročanja, ne da bi poslali preveč opomnikov pokriva, kako to storiti, ne da bi ljudi izčrpal.

Dodaj popust, da izsiliš pretvorbo. To prekrije problem časovnice z maržo. Kupil boš nekaj ponovnih naročil, ki so tako ali tako prihajala, in kupce izučil, da čakajo na kupon. Zdravi simptom.

Pravi problem je, da za vse izračunavaš en interval. Pogostost, s katero kupec ponovno naroči, ne bi smela biti ena konstanta za celotno trgovino — to je jedrna napaka in vredno je prebrati zakaj pogostosti nakupov ne bi smeli izračunavati enako za vsakega kupca, če želiš celotno utemeljitev.

Kaj storiti namesto tega

Fiksni interval zamenjaj z oceno za vsakega kupca posebej. Potrebuješ tri podatke — velikost pakiranja, stopnjo porabe in datum zadnjega naročila — metoda pa je razložena v kako predvideti, kdaj je kupec pripravljen ponovno naročiti potrošni izdelek. Na kratko:

Izhajaj iz velikosti pakiranja, ne iz privzetka trgovine. Kupec večpaka dobi daljšo uro kot kupec enkratnega, samodejno.
Uporabi realno dnevno porabo z embalaže kot prvo domnevo za nove kupce.
Popravi z resničnim vedenjem po drugem naročilu. Kupčev dejanski zadnji interval je najboljši napovedovalec naslednjega — če se je vrnil v 22 dneh, ga nehaj obravnavati kot 30-dnevnega kupca.
Kupce razvrsti po tem, kako hitro dejansko ponovno naročijo, da se tok prilagodi, namesto da se sproža po enem koledarju. Ta razvrstitev je kako segmentirati kupce po dejanskem vedenju ponovnega naročanja.

Popoln ne boš. Za večino ljudi boš približno pravilen namesto natančno pravilen za peščico — in to je veliko izboljšanje.

Kaj avtomatizirati

Sprožilec: naročilo z označenim potrošnim izdelkom.
Časovnica: zamik za vsakega kupca posebej, nastavljen na ocenjeni datum pomanjkanja minus rezerva, ki ga poganjata velikost pakiranja in predhodni interval — nikoli ena trdo zakodirana številka.
Segmenti: vsaj intenzivni proti redkim ponovnim naročnikom, da se zamik in sporočilo premakneta s skupino.
Varovala: omejitev pogostosti in čist izhod ob nakupu, da nihče ni nadlegovan, potem ko je kupil.
Cilj: ponovni nakup, merjen proti kontrolni skupini, ki je še na (ali brez) stari časovnici.

Kako izmeriti razliko

Stopnja ponovnih nakupov, tok proti kontroli. Naslov.
Porazdelitev dejanske časovnice ponovnih nakupov proti tvojim napovedanim datumom — to ti pokaže, ali si sistematično prezgoden ali prepozen, in ti omogoča naravnati rezervo.
Prihodek na prejemnika, da se zmaga pokaže v denarju, ne le v odpiranjih.
Stopnja odjav in pritožb pri pošiljanju — signal utrujenosti. Fiksni tok tukaj običajno teče bolj vroče kot napovedani.

Če že poganjaš fiksni opomnik, je najčistejši test, da rezino novih kupcev premakneš na časovnico za vsakega kupca posebej in primerjaš. Vrzel je običajno očitna v nekaj ciklih.

Kje se umesti Omnisend

Časovnica za vsakega kupca posebej potrebuje orodje, ki zna razvejati glede na oznake izdelkov in zgodovino kupca ter nastaviti zamike, ki se razlikujejo po segmentih. Svoje trgovine sem prestavil na Omnisend, potem ko sem ga preizkusil proti Klaviyu, in razlog, zakaj je tukaj pomemben, je, da je bila izgradnja “razdeli intenzivne proti redkim ponovnim naročnikom, nato vsako vejo časovno odmeri drugače” povleci in spusti namesto izdelava po meri. Orodje, zaklenjeno na en globalen zamik, te rešitve sploh ne more izraziti.

Pošteni del: nobena platforma ne pozna resnične porabe prvega kupca — ve samo za pakiranje, ki ga je kupil, tako da je prvo naročilo vedno izobražena domneva. Natančnost pride z drugim naročilom. In orodje ne bo rešilo opomnika, uperjenega v izdelek, ki ljudem ni bil všeč. Omnisend je partner Shopimation prek pridruženega programa; priporočam ga iz vsakodnevne uporabe, brezplačna raven pa je dovolj, da svoj opomnik ponovno zgradiš kot segmentiran tok za vsakega kupca posebej in ga preizkusiš proti staremu fiksnemu.

Tvoj naslednji korak

Poglej svoj trenutni opomnik za dopolnjevanje in preveri, ali se sproži na en fiksni dan za vse. Če se, si našel svoje puščanje. Izberi en izdelek, nove kupce razdeli po velikosti pakiranja kot prvi grob segment in vsakemu daj drugačen zamik — nato načrtuj natančno različico s kako predvideti, kdaj je kupec pripravljen ponovno naročiti potrošni izdelek.

Subscription and Consumable Products

Why Fixed Replenishment Reminders Fail When Usage Varies

Posted on July 18, 2026 by

A fixed replenishment reminder — “email everyone 30 days after they buy” — fails because your customers don’t use the product at the same speed, so a single interval is right for almost none of them. It’s early for light users and late for heavy ones. The early reminders get ignored and train people to tune you out; the late ones arrive after the customer already ran out and rebought somewhere else. One number can only match the average customer, and in most consumable categories the spread around that average is wide enough that “average” describes a small minority. This piece explains why the fixed model breaks, how much revenue it quietly costs, and what to do instead.

The situation you probably recognise

You set up a reorder reminder because it seemed obvious: people finish the product, so remind them to buy again. You picked a delay — 30 days feels reasonable — and switched it on. It does something. A few reorders trickle in. But the reorder rate is lower than you’d hoped, unsubscribes on that send are creeping up, and you can’t shake the feeling that plenty of customers are running out without ever coming back.

They are. The reminder isn’t broken; the assumption underneath it is. A fixed delay assumes everyone consumes at one rate, and consumables are exactly the category where that assumption falls apart.

Why usage varies more than you’d think

Consider the reasons two people burn through the same product at different speeds:

Household size. One person’s shampoo lasts a household of four a fortnight and a single person two months.
Dose and habit. Two supplement buyers, same bottle — one takes the label dose, one doubles it, one forgets half the time.
Pack size. A customer who bought the value multipack is on a completely different clock from one who bought a single, even though they’re the “same” customer buying the “same” product.
Seasonality. Sunscreen in July and sunscreen in November aren’t consumed at the same rate by the same person. Neither is allergy spray, or electrolyte mix. That pattern deserves its own handling — see handling seasonal changes in product consumption.

Stack these up and the range of real reorder intervals for a single product can easily run from two weeks to two months. A 30-day reminder sits in the middle of that range and is genuinely well-timed for the thin slice of people near the average.

Where the money actually leaks

Here’s the part that’s easy to miss: a fixed reminder fails in two directions at once, and both cost you.

Late for heavy users — lost sales. The customers who go through the product fastest are, by definition, your best consumable customers. They’d reorder most often and spend the most over a year. A reminder that arrives after they’ve run out reaches them once they’ve already found a substitute. You’re losing your highest-value buyers to bad timing, which is the worst possible group to lose.

Early for light users — list fatigue. The reminder that lands while someone still has two-thirds left reads as pestering. They ignore it, then ignore the next one, then unsubscribe or mark it spam. Now your deliverability dips for everyone, and you’ve damaged the channel to save a reminder that was never going to convert that person anyway.

A rough illustration: imagine 1,000 buyers of a product whose real reorder intervals spread from 15 to 60 days. A single day-30 send might land “in the right window” for maybe a couple of hundred of them. The rest get it too early or too late. Even a modest improvement in timing across the other 800 — turning a fraction of missed reorders into actual reorders — usually dwarfs anything you’d get from a nicer template or a discount. The numbers are illustrative, but the asymmetry is real: timing beats creative in this flow.

Why the usual “fixes” don’t fix it

Just send it earlier. Move the reminder to day 20 and you rescue some heavy users but pester even more light users. You’ve shifted the failure, not removed it.

Send more reminders. A reminder at 20, 30, and 45 days covers more of the range, but now light users get three emails they don’t need, and you’ve made the fatigue problem worse. Volume isn’t a substitute for accuracy. If you’re going to send a short series, it has to be paced deliberately — how to build a reorder journey without sending too many reminders covers doing that without wearing people out.

Add a discount to force the conversion. This papers over the timing problem with margin. You’ll buy some reorders that were coming anyway and train customers to wait for the coupon. It treats the symptom.

The real issue is that you’re calculating one interval for everyone. The frequency at which a customer reorders shouldn’t be one store-wide constant — this is the core mistake, and it’s worth reading why purchase frequency should not be calculated the same for every customer if you want the full argument.

What to do instead

Replace the fixed interval with a per-customer estimate. You need three inputs — pack size, a usage rate, and the last order date — and the method is spelled out in how to predict when a customer is ready to reorder a consumable product. The short version:

Start from pack size, not a store default. A multipack buyer gets a longer clock than a single buyer, automatically.
Use a realistic per-day usage from the label as your first guess for new buyers.
Correct with real behaviour after order two. A customer’s actual last interval is the best predictor of their next one — if they came back in 22 days, stop treating them like a 30-day customer.
Group customers by how fast they actually reorder so the flow adapts instead of firing on one calendar. That grouping is how to segment customers by real reorder behaviour.

You won’t be perfect. You’ll be roughly right for most people instead of precisely right for a handful — and that’s a large improvement.

What to automate

Trigger: an order with a tagged consumable.
Timing: a per-customer delay set to estimated run-out minus a buffer, driven by pack size and prior interval — never one hard-coded number.
Segments: at minimum, heavy vs. light reorderers, so the delay and message shift with the group.
Guardrails: a frequency cap and a clean exit-on-purchase so nobody gets nagged after they’ve bought.
Goal: the reorder, measured against a control group still on (or off) the old timing.

How to measure the difference

Reorder rate, flow vs. control. The headline.
Distribution of actual reorder timing against your predicted dates — this shows you whether you’re systematically early or late and lets you tune the buffer.
Revenue per recipient, so the win shows up in money rather than only in opens.
Unsubscribe and complaint rate on the send — the fatigue signal. A fixed-interval flow usually runs hotter here than a predicted one.

If you already run a fixed reminder, the cleanest test is to move a slice of new buyers onto per-customer timing and compare. The gap is usually obvious within a couple of cycles.

Where Omnisend fits

Per-customer timing needs a tool that can branch on product tags and customer history and set delays that differ by segment. I switched my own stores to Omnisend after testing it against Klaviyo, and the reason it matters here is that building “split heavy vs. light reorderers, then time each branch differently” was drag-and-drop rather than a custom build. A tool locked to one global delay can’t express the fix at all.

The honest part: no platform knows a first-time buyer’s real usage — it only knows the pack they bought, so order one is always an educated guess. Accuracy arrives with the second order. And a tool won’t save a reminder pointed at a product people didn’t like. Omnisend is an affiliate partner of Shopimation; I recommend it from daily use, and the free tier is enough to rebuild your reminder as a segmented, per-customer flow and test it against the old fixed one.

Your next step

Look at your current replenishment reminder and check whether it fires on one fixed day for everyone. If it does, you’ve found your leak. Pick one product, split new buyers by pack size as a first crude segment, and give each a different delay — then plan the accurate version with how to predict when a customer is ready to reorder a consumable product.