Povečanje povprečne vrednosti naročila

Kaj priporočiti, ko izdelek nima očitnega dodatka

Posted on July 18, 2026 by

Ko izdelek nima očitnega dodatka, nehajte iskati fizičen priključek in raje priporočajte glede na cilj. Večina posameznih izdelkov še vedno spada v neko opravilo, ki ga kupec želi dokončati, v okus, ki ga je pravkar razkril, ali v rutino, ki jo bo ponavljal. Zato odgovor le redko glasi “ni česa pripeti”. Ponavadi je “pripenjate na napačni osi”. Priporočite naslednji korak v istem opravilu, drug izdelek iz iste slogovne skupine ali potrošni material, ki ohranja prvi nakup uporaben. Ovitek za telefon ima očiten dodatek. Stenska ura, dišeča sveča, volnena odeja — ti ga nimajo, in prav tu trgovci obupajo ter pokažejo splošno mrežo najbolje prodajanih izdelkov. Ta stran govori o tem, kaj tja postaviti namesto tega.

“Ni očitnega dodatka” je skoraj vedno težava zornega kota

Razmišljanje o dodatkih je namerno ozko. Sprašuje: kaj se fizično pritrdi na ta predmet? Fotoaparat sprejme objektiv, kolo ključavnico, prenosnik torbo. Marsikateri katalog je poln izdelkov, pri katerih to vprašanje ne vrne ničesar. Uokvirjen tisk. Posamezna skodelica. Svilena ruta. Kozarec medu.

Napaka je, da predmet obravnavamo kot enoto. Kupec ni kupil predmeta v abstraktnem smislu — kupil ga je, da bi nekaj naredil, nekaj občutil ali nekaj podaril. Priporočilo zastavite okoli tega in zid “brez dodatka” izgine. Ruta je del kompleta oblačil. Med je del zajtrka ali darila. Tisk je del stene, ki ima še druga prazna mesta.

Ta premik je pomemben, ker je alternativa — vrsta vaših najbolje prodajanih izdelkov — najšibkejše, kar lahko pokažete. Prezre vse, kar ste pravkar izvedeli o tem konkretnem kupcu. Če želite daljšo razlago, zakaj sorodno videti ponudbe vseeno zgrešijo, jo pokrije zakaj večina ponudb navzkrižne prodaje odpove, tudi ko so izdelki povezani. Tukaj ostajam pri praktičnem vprašanju: kaj torej dejansko priporočiti?

Priporočajte po eni od štirih osi

Ko je os dodatkov prazna, štiri druge ponavadi niso. Izberite tisto, ki izdelku najbolje ustreza.

Isto opravilo. Kaj kupec poskuša dokončati in kaj je naslednji del tega? Nekdo, ki kupi en sam jedilni stol, morda opremlja sobo, ne samo enega mesta. Nekdo, ki kupi podlogo za jogo, začenja vadbo, ki potrebuje tudi blok in trak. Podloga v strogem pomenu nima “dodatka”, ima pa očiten naslednji korak. Vprašajte se, kaj po kupčevi glavi sledi temu nakupu, ne kaj se zaskoči na škatlo.

Isti okus. Nakup je signal o slogu prav toliko kot o potrebi. Kupec, ki izbere umirjeno, zemeljsko keramično skledo, vam pove, katere druge vaše kose bo imel rad. Tu sploh ne potrebujete funkcionalne povezave — potrebujete skladno. Prav tu zmaga trgovina s stotinami podobnih nadomestkov, če so priporočila skrbno izbrana in ne naključna; kako izgleda dobro priporočilo izdelka pri velikem katalogu se v to težavo poglobi.

Potrošni material ali nega. Nekateri samostojni izdelki tiho potrebujejo vzdrževanje. Usnjeni čevlji potrebujejo zaščitno sredstvo. Litoželezna ponev potrebujejo olje za utrjevanje. Volnen plašč potrebuje strojček za odstranjevanje muck in vreče za shranjevanje. Nič od tega ni dodatek v smislu priključka, a vse ohranja prvi nakup delujoč in vse je naravna, poštena stvar, ki jo lahko ponudite.

Kontekst darila. Precejšen delež izdelkov “brez dodatka” kupijo za darila. Sveča, lepa beležnica, steklenica nečesa. Če se naročilo obnaša kot darilo — en izdelek, dodano darilno sporočilo, dostava na drug naslov — je uporaben dodatek embalaža, voščilnica ali drug majhen izdelek, ki darilo zaokroži.

Vseh štirih ne uporabite pri vsem. Izberete tisto, ki se prilega, in ostale spustite.

Naj podatki poimenujejo pare, ki jih sami ne bi uganili

Vaša lastna zgodovina naročil je tu najbolj poštén vir in redno prekaša intuicijo. Poglejte, kaj dejansko konča v istem košariku ali v drugem naročilu v nekaj tednih pri ljudeh, ki so kupili “nezdružljiv” izdelek. Včasih je povezava nevidna, dokler je ne pokažejo podatki — kupci, ki kupijo vašo sveče brez vonja, pogosto tudi, čudno pogosto, tri tedne pozneje kupijo medeninast dušilnik.

To je vzorec, na katerem lahko gradite, in ni tak, do katerega bi se dokopali z razmišljanjem s strani izdelka. Metoda za iskanje teh je poglavje zase: kako najti kombinacije izdelkov, ki jih kupci naravno kupujejo skupaj. Začnite tam, preden ročno ugibate pare za velik katalog, ker se ugibanje slabo prilagaja obsegu, podatki pa ne.

Kje dejansko odteka prihodek

Izguba je tu tiha, zato ostaja nepopravljena. Trgovina z nekaj tisoč naročili na mesec, kjer je morda 40 % naročil posameznih “težko združljivih” izdelkov, tem kupcem prikazuje bodisi nič bodisi zastarelo vrsto najbolje prodajanih izdelkov. Recimo, da je povprečno naročilo 35 € in resnično ustrezen drugi predlog dvigne celo 8 % teh kupcev posameznega izdelka za povprečno 14 €. (Ponazoritveno — vaše številke bodo drugačne.) Pri tisoč takih naročilih na mesec je to več kot 1.100 € dodatnega prihodka, ki ste ga puščali na mizi, od kupcev, ki so že zaključevali nakup. Brez novih stroškov za oglase. Brez novega prometa. Le boljši odgovor na vprašanje “kaj še?”.

Odtekanje ni v tem, da ljudje rečejo ne. V tem je, da nikoli niste podali ustrezne ponudbe, zato ni bilo ničesar, čemur bi rekli da.

Primer iz trgovine

Vzemimo majhno trgovino z izdelki za dom. Njihov najbolje prodajan izdelek je ročno vlita sveča za 29 € v enem vonju in dve leti je e-poštno sporočilo po nakupu prikazovalo “naši najboljši izbori” — istih štiri izdelke vsem. Prihodek od navzkrižne prodaje kupcem sveč je bil skoraj ničen.

Spremenili so eno stvar. Namesto najboljših izborov so kupci sveč dobili tri predloge, zgrajene na osi okusa: dva druga vonja iz iste tonske družine in preprosto keramično stojalo, ki se je ujemalo z glazuro sveče. Besedilo e-pošte je nehalo govoriti “morda vam bo všeč tudi” in začelo govoriti “če vam je bila všeč [vonj], so ti trije po občutku najbližji”. To je celotna sprememba — isto orodje, isto pošiljanje, drugačna logika. (Ponazoritveni primer; prilagodite pristop svojemu katalogu.)

Bistvo ni v natančnem dvigu. Bistvo je, da sveča nikoli ni bila resnično “nezdružljiva”. Le pripenjati jo je bilo treba po okusu namesto po funkciji, in dodati vrstico besedila, ki je priznala, kaj počne.

Kaj avtomatizirati

Zorni kot spremenite v ponovljiv potek namesto v enkratno merchandising opravilo.

Sprožilec: nakup izdelka z oznako “brez naravnega dodatka” ali brskanje/dodajanje takega izdelka v košarico.
Segment: razdelite po osi, ki ustreza izdelku. Naročila, ki se obnašajo kot darilo, dobijo darilne dodatke; kupci rednega potrošnega materiala dobijo izdelek za nego; kupci, ki jih vodi slog, dobijo skupino po okusu.
Časovnica: za nego in potrošni material se sporočilo po nakupu nekaj dni po dostavi obnese bolje kot ob zaključku nakupa, ker kupec zdaj stvar poseduje in čuti njeno vzdrževanje. Za izbore po okusu deluje ob zaključku nakupa ali takoj po njem.
Kanal: e-pošta za premišljene predloge, ki jih je mogoče prebrskati; SMS le, če je ponudba resnično časovno občutljiva.
Vsebina: dva do štiri izbori, en jasen razlog, zakaj so prikazani, en gumb. Ne zid.
Cilj: ustrezen drugi nakup, merjen kot stopnja pripetja, ne surovi kliki.

Če vam nagon veleva, da bi te izdelke raje združili v pakete kot prikazali ločeno, je to prava razpotje — paketi ali posamezna priporočila: kaj bolj dvigne vrednost naročila to pravilno pretehta.

Kako izmeriti, ali deluje

Spremljajte tri številke, ne deset:

Stopnjo pripetja pri prej neparjenih izdelkih — delež teh naročil, ki dodajo priporočen izdelek. To je neposredna tablica z rezultatom.
Povprečno vrednost naročila za kupce teh izdelkov, pred in po tem, ko preklopite z najbolje prodajanih na izbore po osi.
Klikovnost priporočil kot diagnostiko. Veliko klikov, a nizko pripetje pomeni, da je izbor zanimiv, a napačen; malo klikov pomeni, da zorni kot ali besedilo ne prime.

Če se nič ne premakne, je poštena razlaga ponavadi, da je os za ta izdelek napačna, ne da je izdelek nezdružljiv. Poskusite z drugo od štirih.

Kje se vklopi Omnisend

Vsaka spodobna e-poštna platforma zna poslati sporočilo po nakupu. Kar je to v mojih trgovinah olajšalo, so bile Omnisendove oznake izdelkov in bloki dinamičnih priporočil — lahko označim nabor “težko združljivih” SKU-jev, vsakega povežem s skupino po okusu ali izdelkom za nego in pustim, da se blok napolni za vsakega kupca posebej, namesto da vzdržujem ročno izbrane sezname, ki zastarijo v trenutku, ko se zaloga spremeni. K njemu sem prišel, potem ko sem ga preizkusil proti Klaviyu, večinoma zaradi bolj samopostrežne nastavitve in združene e-pošte in SMS-a v enem poteku.

Pošten pridržek: orodje napolni blok, a ne more odločiti, kateri od štirih osi izdelek pripada. Ta presoja je vaša in je del, ki dejansko prisluži dodatno naročilo. Omnisend je partner Shopimationa v pridruženem programu; priporočam ga iz vsakodnevne uporabe, brezplačna raven pa zadošča, da najprej preizkusite priporočila enega izdelka na majhnem segmentu.

Vaš naslednji korak

Izberite svoj eno najslabši “nezdružljiv” izdelek — tistega, ki v e-poštnem sporočilu po nakupu trenutno prikazuje splošne najbolje prodajane izdelke. Odločite se, kateri od štirih osi resnično pripada, povežite mu tri pristne izbore in prepišite naslovno vrstico, da poimenuje razlog. Nato mesec dni spremljajte njegovo stopnjo pripetja. Ko enkrat deluje en izdelek, se ista metoda razširi po katalogu — in spreminjanje tega posameznega izdelka v polnejšo košarico je projekt zase, vreden truda: kako naročilo z enim izdelkom spremeniti v celovito rešitev.

Average Order Value Growth

What to Recommend When a Product Has No Obvious Accessory

Posted on July 18, 2026 by

When a product has no obvious accessory, stop hunting for a physical add-on and recommend by outcome instead. Most single items still sit inside a job the customer is trying to finish, a taste they’ve just revealed, or a routine they’ll repeat. So the answer is rarely “there’s nothing to pair.” It’s usually “you’re pairing on the wrong axis.” Recommend the next step in the same job, another item from the same style cluster, or the consumable that keeps the first purchase useful. A phone case has an obvious accessory. A wall clock, a scented candle, a wool blanket — those don’t, and that’s exactly where merchants give up and show a generic bestseller grid. This page is about what to put there instead.

“No obvious accessory” is almost always a framing problem

Accessory thinking is narrow on purpose. It asks: what physically attaches to this object? A camera takes a lens, a bike takes a lock, a laptop takes a sleeve. Plenty of catalogs are full of products where that question returns nothing. A framed print. A single mug. A silk scarf. A jar of honey.

The mistake is treating the object as the unit. The customer didn’t buy an object in the abstract — they bought it to do something, to feel something, or to give something. Reframe the recommendation around that, and the “no accessory” wall disappears. The scarf is part of an outfit. The honey is part of a breakfast, or a gift. The print is part of a wall that has other empty spots on it.

That shift matters because the alternative — a row of your top sellers — is the weakest thing you can show. It ignores everything you just learned about this specific buyer. If you want the longer argument on why related-looking offers still miss, why most cross-sell offers fail even when the products are related covers it. Here I’ll stay on the practical question: what do you actually recommend?

Recommend along one of four axes

When the accessory axis is empty, four others usually aren’t. Pick whichever the product fits best.

The same job. What is the customer trying to finish, and what’s the next piece of it? Someone buying a single dining chair may be furnishing a room, not a single spot. Someone buying a yoga mat is starting a practice that also wants a block and a strap. The mat has no “accessory” in the strict sense, but it has an obvious next step. Ask what comes after this purchase in the customer’s head, not what clips onto the box.

The same taste. A purchase is a signal about style as much as about need. A buyer who picks a muted, earthy ceramic bowl is telling you which of your other pieces they’ll like. You don’t need a functional link at all here — you need a coherent one. This is where a store with hundreds of near-substitutes wins, if the recommendations are curated rather than random; what a good product recommendation looks like for a large catalog goes deeper on that problem.

The consumable or the care. Some standalone products quietly need upkeep. Leather shoes want protectant. A cast-iron pan wants seasoning oil. A wool coat wants a fabric shaver and storage bags. None of these is an accessory in the plug-in sense, but each keeps the first purchase working, and each is a natural, honest thing to offer.

The gift context. A meaningful share of “no accessory” products are bought as presents. A candle, a nice notebook, a bottle of something. If the order behaves like a gift — single item, gift note added, shipping to a different address — the useful add-on is wrapping, a card, or a second small item to round the gift out.

You don’t apply all four to everything. You choose the one that fits, and you let the rest go.

Let the data name the pairs you’d never guess

Your own order history is the most honest source here, and it routinely beats intuition. Look at what actually ends up in the same basket, or in a second order within a few weeks, for people who bought the “unpairable” product. Sometimes the connection is invisible until the data shows it — the customers who buy your unscented candle also, oddly often, buy the brass snuffer three weeks later.

That’s a pattern you can build on, and it’s not one you’d have reasoned your way to from the product page. The method for finding these is its own topic: how to find the product combinations customers naturally buy together. Start there before you hand-guess pairings for a large catalog, because guessing scales badly and the data doesn’t.

Where the revenue actually leaks

The loss here is quiet, which is why it goes unfixed. A store with a few thousand orders a month, where maybe 40% of orders are single “hard-to-pair” items, is showing those buyers either nothing or a stale bestseller row. Say the average order is €35 and a genuinely relevant second suggestion lifts even 8% of those single-item buyers by an average of €14. (Illustrative — your numbers will differ.) On a thousand such orders a month that’s over €1,100 in added revenue you were leaving on the table, from customers who were already checking out. No new ad spend. No new traffic. Just a better answer to “what else?”

The leak isn’t that people say no. It’s that you never made a relevant offer, so there was nothing to say yes to.

A store example

Take a small home-goods shop. Their best seller is a €29 hand-poured candle in one scent, and for two years the post-purchase email showed “our top picks” — the same four products to everyone. Cross-sell revenue from candle buyers was close to nothing.

They changed one thing. Instead of top picks, candle buyers got three suggestions built on the taste axis: two other scents in the same tonal family and a simple ceramic holder that matched the candle’s glaze. The email copy stopped saying “you might also like” and started saying “if you liked the [scent], these three are the closest in feel.” That’s the whole change — same tool, same send, different logic. (Illustrative example; adapt the framing to your own catalog.)

The point isn’t the exact lift. It’s that the candle was never truly “unpairable.” It just needed pairing on taste instead of function, and a line of copy that admitted what it was doing.

What to automate

Turn the framing into a repeatable flow rather than a one-off merchandising chore.

Trigger: purchase of a product tagged “no natural accessory,” or browse/add-to-cart of one.
Segment: split by the axis that fits the product. Gift-behaving orders get gift add-ons; routine consumable buyers get the care item; style-led buyers get the taste cluster.
Timing: for care and consumables, a post-purchase message a few days after delivery lands better than at checkout, because the customer now owns the thing and feels its upkeep. For taste-based picks, at checkout or immediately after works.
Channel: email for the considered, browsable suggestions; SMS only if the offer is genuinely time-sensitive.
Content: two to four picks, one clear reason they’re shown, one button. Not a wall.
Goal: a relevant second purchase, measured as attach rate, not raw clicks.

If your instinct is to bundle these instead of showing them separately, that’s a real fork in the road — bundles or individual recommendations: which raises order value more weighs it properly.

How to measure whether it’s working

Watch three numbers, not ten:

Attach rate on the previously-unpaired products — the share of those orders that add a recommended item. This is the direct scoreboard.
Average order value for buyers of those products, before and after you switch from bestsellers to axis-based picks.
Recommendation click-through, as a diagnostic. High clicks but low attach means the pick is interesting but wrong; low clicks means the framing or copy isn’t landing.

If nothing moves, the honest read is usually that the axis is wrong for that product, not that the product is unpairable. Try a different one of the four.

Where Omnisend fits

Any decent email platform can send a post-purchase message. What made this easy in my own stores was Omnisend’s product-tagging and dynamic recommendation blocks — I can tag a set of “hard-to-pair” SKUs, wire each to a taste cluster or a care item, and let the block populate per customer instead of maintaining hand-picked lists that go stale the moment stock changes. I landed on it after testing it against Klaviyo, mostly for the more self-serve setup and the combined email/SMS in one flow.

Honest caveat: the tool populates the block, but it can’t decide which of the four axes a product belongs to. That judgment is yours, and it’s the part that actually earns the extra order. Omnisend is an affiliate partner of Shopimation; I recommend it from daily use, and the free tier is enough to test one product’s recommendations on a small segment first.

Your next step

Pick your single worst “unpairable” product — the one whose post-purchase email currently shows generic bestsellers. Decide which of the four axes it really belongs to, wire three genuine picks to it, and rewrite the header line to name the reason. Then watch its attach rate for a month. Once one product works, the same method scales across the catalog — and turning that single item into a fuller basket is its own worthwhile project: how to turn a single-item order into a complete solution.