Dostavljivost in higiena email baze

Kako izboljšati ugled pošiljatelja po neuspešni kampanji

Posted on July 18, 2026 by

Zaporedje popravila je: ustavi, skrči, popravi, obnovi. Ustavite koledar kampanj, pustite delovati le svoje najbolj angažirane avtomatizacije in ugotovite natanko, kaj je slaba kampanja povzročila — skok prijav, skok zavrnitev ali oboje. Umaknite segment, ki ga je povzročil. Nato obnovite zaupanje tako, da nekaj tednov pošiljate le svojim najbolj angažiranim naročnikom in občinstvo postopoma širite, tako kot bi ogrevali novo domeno, medtem pa spremljate svoj ugled v Google Postmaster Tools, dokler si ne opomore. Ponudniki poštnih predalov močno ocenjujejo nedavno vedenje, tako da obdobje čistega, dobro sprejetega pošiljanja škodo resnično popravi — a to zahteva tedne discipline, ne prijave na podporo. Ta članek predpostavlja, da veste, katera kampanja vas je prizadela. Če so vam odpiranja strmoglavila in ne veste, zakaj, začnite s kako si opomoči po nenadnem padcu odpiranj emailov in se vrnite, ko boste našli vzrok.

Kaj je “slaba kampanja” običajno bila

V praksi je kampanja, ki uniči ugled trgovine, skoraj vedno ena s kratkega seznama. Zasip na segment, ki ni bil pošiljan eno leto — pogosto za črni petek ali razprodajo. Uvožen seznam, v katerega nikoli ni bilo pravilno privoljeno: kontakti s sejma, kupljena datoteka, izvoz iz prejšnjega posla. Pošiljanje na celoten seznam, ko vaše običajne kampanje potujejo na angažirani del, tako da je naenkrat zadelo tisoče napol mrtvih naslovov. Ali, redkeje, tehnična odpoved — sprememba DNS, ki je sredi kampanje pokvarila vašo avtentikacijo, tako da je pošta odhajala z neuspešnim SPF ali DKIM.

Simptomi se pokažejo hitro. Stopnja prijav skoči pri tistem enem pošiljanju. Trde zavrnitve poskočijo, če je bilo občinstvo zastarelo. Nato posledica: naslednja pošiljanja, na povsem dobra občinstva, delujejo slabše — odpiranja padejo, več pošte gre v neželeno — ker se je kazen pripela na vašo domeno, ne na kampanjo.

Zakaj čakanje ne deluje — in tudi nadaljevanje ne

Po slabem pošiljanju se pojavita dva nagona in oba sta napačna.

Prvi je, da ne spremenite ničesar in upate, da bo minilo. Ne bo, ker je vaš ugled tekoča ocena: vsako dodatno pošiljanje na isto vrsto občinstva dodaja sveže negativne dokaze. Trgovina, ki je v petek zasula svoj hladen seznam in nato v torek požene svojo običajno kampanjo na isti seznam, si znova škoduje dvakrat na teden.

Drugi nagon je, da za mesec dni popolnoma utihnete. Molk ustavi krvavitev, a ničesar ne zaceli — ponudniki vas sodijo po pošti, ki jo pošiljate, in nič pošte pomeni nič novih pozitivnih signalov, ki bi odtehtali slabe. Vrnete se mesec pozneje z isto poškodovano oceno in, kot dodatek, s hladnejšim seznamom.

Okrevanje potrebuje tretjo možnost: pošiljajte še naprej, a le pošto, ki bo dokazljivo dobrodošla. Majhna občinstva, visoka angažiranost, čisti rezultati. Gradite nov zapis dosežkov na vrhu slabe podatkovne točke, dokler slaba podatkovna točka ne postane nereprezentativna.

Načrt okrevanja, teden za tednom

Spodnjo časovnico obravnavajte kot predlogo za prilagoditev, ne kot urnik z magičnimi številkami — koliko časa vzame vsaka faza, je odvisno od tega, kako huda je škoda in kako pogosto pošiljate.

1.–2. dan: ustavi in diagnosticiraj. Ustavite vse načrtovane kampanje. Transakcijske emaile pustite pri miru (računi in obvestila o pošiljanju morajo iti ven in so običajno vaša najbolje sprejeta pošta). Pustite delovati avtomatizacije z visoko namero — zapuščena košarica, pozdravni tok — saj potujejo k ljudem, ki se prav zdaj aktivno ukvarjajo z vašo trgovino, kar je natanko signal, ki ga želite. Nato odprite poročilo slabe kampanje in poimenujte odpoved: kateri segment, prijave ali zavrnitve, kako daleč zunaj vaših običajnih številk. Med tem preverite svoje SPF, DKIM in DMARC, da izključite tehnično različico.

2.–3. dan: karantena vzroka. Umaknite občinstvo, ki je ustvarilo skok — star segment, uvoz, nikoli angažiran rep. Ne nujno za vedno, a za celotno obdobje okrevanja. Če je bil temeljni vzrok seznam brez privolitve, ga umaknite trajno; te pošte ni mogoče narediti dobrodošle. Preglejte tudi pošiljanja z veliko prijavami za vzorec, ki ga lahko popravite, saj se ugled, popravljen ob preživelem vzroku, spet pokvari — kako zmanjšati prijave marketinških emailov kot neželeno pošto pokriva popravke višje ob toku.

1.–2. teden: pošiljaj majhno in odlično. Zgradite segment za okrevanje iz svojih resnično angažiranih kontaktov — ljudi, ki so nedavno kliknili ali kupili, recimo v zadnjih 30–60 dneh. Natančen recept je v segmentiranje angažiranih in neaktivnih naročnikov. Pošiljajte le njim, s svojo običajno pogostostjo ali malce pod njo, s svojo najboljšo vsebino: stvarmi, ki jih ta skupina zgodovinsko odpira in klika. Cilj teh pošiljanj ni prihodek. Je niz kampanj z močno angažiranostjo in skoraj ničelnimi prijavami, ki pristajajo pred ponudniki, ki so trenutno sumničavi do vas.

3.–6. teden: širi postopoma. Ko Postmaster Tools pokaže, da je vaša stopnja neželene pošte spet blizu izhodišča in vaša angažirana pošiljanja izgledajo normalno, občinstvo razširjajte po korakih — 90-dnevna angažiranost, nato 120, pri čemer spremljate številke vsake razširitve pred naslednjo. Če kak korak prinese prijave ali uvrstitev v neželeno pošto nazaj, se umaknite en korak in dlje počakajte. Ta stopnjevana razširitev je ista mišica kot ogrevanje sveže infrastrukture, kako ogreti novo domeno za pošiljanje spletne trgovine pa podrobneje opisuje logiko ritma. Nikoli angažirani kontakti, ki ste jih dali v karanteno, se v kampanje sploh ne vrnejo — gredo skozi pravi sunset postopek ali s seznama.

Ena poštena pridržka: če je bila škoda dovolj huda — običajno zasip s kupljenim seznamom, ki je zadel pasti za neželeno pošto — lahko okrevanje traja dlje od tega, v najhujših primerih pa se trgovine na koncu preselijo na svežo poddomeno in jo ogrejejo od nič. To je zadnja možnost, ne načrt. Najprej poskusite popravilo.

Kaj to stane, konkretno

Ponazoritvene številke, da naredijo kompromis viden. Trgovina, ki običajno pošilja 15.000 kontaktom, bi lahko tri tedne pošiljala le 4.000 med okrevanjem. Če email tej trgovini prinaša, recimo, 8.000 € na mesec, bi skrčeno občinstvo lahko stalo nekaj tisoč evrov kratkoročnega prihodka kampanj. Proti temu: poškodovana domena tiho meče med neželeno pošto tudi vaše tokove za zapuščeno košarico in pozdrav, ti pa pogosto zaslužijo več na sporočilo kot kampanje. Trije disciplinirani tedni so cenejši od mesecev vsake avtomatizacije, ki dostavlja premalo. (Vse številke so ponazoritvene — vstavite svoje.)

Kako veste, da deluje

Google Postmaster Tools je glavni inštrument: spremljajte, kako graf stopnje neželene pošte pade nazaj pod 0,1–0,3 % in kako ugled domene splezá iz jame. Nastavitev in razlaga sta obravnavani v kako spremljati ugled domene za email spletne trgovine.
Stopnja klikov angažiranega segmenta, ki se vrne k svoji normi izpred incidenta, vam pove, da si uvrstitev v mapo “prejeto” opomore — vedenje te skupine se ni spremenilo, torej če so njihove številke padle in nato okrevale, je bila to uvrstitev.
Stopnji prijav in zavrnitev na kampanjo naj bosta na ali pod vašim starim izhodiščem pri vsakem pošiljanju med okrevanjem. Eno čisto pošiljanje ne dokazuje ničesar; šest zapored je trend.
Prihodek avtomatizacij — tokovi za košarico in pozdrav — ki se vrne na normalno, je znak, da je popravilo doseglo pošto, ki vam prinaša denar.

Kje se vključi Omnisend

Načrt okrevanja je nevtralen do platforme; kar od svojega orodja potrebujete, je zmožnost, da ga izvedete natančno. V Omnisendu — ki ga v svojih trgovinah uporabljam po tem, ko sem ob njem preizkusil Klaviyo — so delovni deli segmenti po nedavnosti angažiranosti za gradnjo občinstva za okrevanje in vsakega obroča razširitve, seznami za umik za karanteno segmenta, ki je povzročil škodo, poročila na kampanjo, ki prikažejo prijave in zavrnitve na dan pošiljanja, ter nadzor na ravni toka, tako da avtomatizacije tečejo naprej, medtem ko kampanje mirujejo. Shopimation je partner Omnisenda, tako da to upoštevajte, kot želite; razlog, da ga priporočam, je, da je prav ta vrsta discipline pri segmentih enostavna za nastavitev brez agencije. Za tedne dobrega vedenja pa ne obstaja nobena platformna bližnjica. Ta del si zaslužite.

Vaš naslednji korak

Danes: ustavite koledar, odprite poročilo slabe kampanje in zapišite tri številke — njeno stopnjo prijav, njeno stopnjo zavrnitev in segment, na katerega je šla. To vam pove, kaj umakniti in kako previdna mora biti vaša obnova. Nato nastavite spremljanje ugleda domene pred svojim prvim pošiljanjem med okrevanjem, da krmarite po inštrumentih namesto po občutku.

Deliverability and List Hygiene

How to Improve Sender Reputation After a Bad Campaign

Posted on July 17, 2026 by

The repair sequence is: stop, shrink, fix, rebuild. Pause your campaign calendar, keep only your highest-engagement automations running, and work out exactly what the bad campaign did — complaint spike, bounce spike, or both. Suppress the segment that caused it. Then rebuild trust by sending only to your most engaged subscribers for a couple of weeks and widening the audience gradually, the way you’d warm up a new domain, while watching your reputation in Google Postmaster Tools until it recovers. Mailbox providers score recent behavior heavily, so a stretch of clean, well-received sending genuinely does repair the damage — but it takes weeks of discipline, not a support ticket. This article assumes you know which campaign hurt you. If your opens fell off a cliff and you don’t know why, start with how to recover from a sudden drop in email opens and come back once you’ve found the cause.

What a “bad campaign” usually was

In practice, the campaign that wrecks a store’s reputation is almost always one of a short list. A blast to a segment that hadn’t been mailed in a year — often for Black Friday or a clearance push. An imported list that was never properly consented: trade-show contacts, a purchased file, an export from a previous business. A send to the full list when your normal campaigns go to the engaged slice, so thousands of half-dead addresses got hit at once. Or, less often, a technical failure — a DNS change that broke your authentication mid-campaign, so mail went out failing SPF or DKIM.

The symptoms show up fast. Complaint rate jumps on that one send. Hard bounces spike if the audience was stale. Then the knock-on effect: the next sends, to perfectly good audiences, underperform — opens down, more mail in spam — because the penalty attached to your domain, not to the campaign.

Why waiting it out doesn’t work — and neither does carrying on

Two instincts show up after a bad send, and both are wrong.

The first is to change nothing and hope it blows over. It won’t, because your reputation is a running score: every additional send to the same kind of audience adds fresh negative evidence. The store that blasted its cold list on Friday and runs its normal Tuesday campaign to the same list is re-injuring itself twice a week.

The second instinct is to go completely dark for a month. Silence stops the bleeding, but it doesn’t heal anything — providers judge you on the mail you send, and no mail means no new positive signals to outweigh the bad ones. You come back a month later with the same damaged score and, as a bonus, a colder list.

Recovery needs a third option: keep sending, but only mail that will be demonstrably welcomed. Small audiences, high engagement, clean results. You’re building a new track record on top of the bad data point until the bad data point stops being representative.

The recovery plan, week by week

Treat the timeline below as a template to adapt, not a schedule with magic numbers — how long each stage takes depends on how bad the damage is and how often you send.

Days 1–2: stop and diagnose. Pause all scheduled campaigns. Leave transactional email alone (receipts and shipping notices must go out, and they’re usually your best-received mail). Keep high-intent automations — abandoned cart, welcome — running; they go to people actively engaging with your store right now, which is exactly the signal you want. Then open the bad campaign’s report and name the failure: which segment, complaints or bounces, how far outside your normal numbers. Check your SPF, DKIM, and DMARC while you’re at it to rule out the technical version.

Days 2–3: quarantine the cause. Suppress the audience that produced the spike — the old segment, the import, the never-engaged tail. Not forever, necessarily, but for the whole recovery period. If the root cause was an unconsented list, suppress it permanently; that mail can’t be made welcome. Also review complaint-heavy sends for a pattern you can fix, since a reputation repaired while the cause survives just breaks again — how to reduce spam complaints from marketing emails covers the upstream fixes.

Weeks 1–2: send small and excellent. Build a recovery segment of your genuinely engaged contacts — people who clicked or bought recently, say within 30–60 days. The precise recipe is in segmenting engaged and unengaged subscribers. Mail only them, at your normal frequency or slightly below, with your best content: things this group historically opens and clicks. The goal of these sends is not revenue. It’s a string of campaigns with strong engagement and near-zero complaints landing in front of the providers that are currently suspicious of you.

Weeks 3–6: widen gradually. Once Postmaster Tools shows your spam rate back near baseline and your engaged sends look normal, expand the audience stepwise — 90-day engagement, then 120, watching each expansion’s numbers before the next. If a step brings complaints or spam placement back, retreat one step and hold longer. This staged expansion is the same muscle as warming up fresh infrastructure, and how to warm up a new ecommerce sending domain describes the pacing logic in more detail. The never-engaged contacts you quarantined don’t rejoin campaigns at all — they go through a proper sunset process or off the list.

One honest caveat: if the damage was severe enough — typically a purchased-list blast that hit spam traps — recovery can take longer than this, and in the worst cases stores end up moving to a fresh subdomain and warming it from zero. That’s the last resort, not the plan. Try the repair first.

What this costs, concretely

Illustrative numbers to make the trade-off visible. A store that normally emails 15,000 contacts might spend three weeks mailing only 4,000 during recovery. If email drives, say, €8,000 a month for that store, the shrunken audience might cost it a couple of thousand euros of short-term campaign revenue. Against that: a damaged domain quietly junks your abandoned-cart and welcome flows too, and those often out-earn campaigns per message. The three disciplined weeks are cheaper than months of every automation underdelivering. (All figures illustrative — plug in your own.)

How to know it’s working

Google Postmaster Tools is the primary instrument: watch the spam-rate graph fall back under 0.1–0.3% and domain reputation climb out of the hole. Setup and interpretation are covered in how to monitor domain reputation for ecommerce email.
Engaged-segment click rate returning to its pre-incident norm tells you inbox placement is recovering — this group’s behavior didn’t change, so if their numbers fell and then recover, that was placement.
Complaint and bounce rates per campaign should sit at or below your old baseline on every recovery send. One clean send proves nothing; six in a row is a trend.
Automation revenue — cart and welcome flows — recovering to normal is the sign the repair reached the mail that makes you money.

Where Omnisend fits

The recovery plan is platform-neutral; what you need from your tool is the ability to execute it precisely. In Omnisend — which I run in my own stores after trying Klaviyo alongside it — the working parts are engagement-recency segments for building the recovery audience and each expansion ring, suppression lists for quarantining the segment that caused the damage, per-campaign reports that show complaints and bounces the day of the send, and flow-level control so automations keep running while campaigns pause. Shopimation is an Omnisend affiliate, so weigh that as you like; the reason I recommend it is that this exact kind of segment discipline is easy to set up without an agency. No platform shortcut exists for the weeks of good behavior, though. That part you earn.

Your next step

Today: pause the calendar, open the bad campaign’s report, and write down three numbers — its complaint rate, its bounce rate, and the segment it went to. That tells you what to suppress and how cautious your rebuild needs to be. Then set up domain reputation monitoring before your first recovery send, so you’re steering by instruments instead of by feel.