NoFraud (Wyllo) vs SOptim — what each app actually blocks

The 30-second version

NoFraud — now listed as Wyllo on the App Store, vendor Wyllo, LLC — screens orders after they're placed. Every order is sent to Wyllo's ML pipeline and, for higher-value cases, a human review team. Wyllo returns an approve or decline. On approved orders, Wyllo offers a chargeback guarantee on its paid tier — if an approved order turns out to be fraudulent and produces a chargeback, Wyllo reimburses the merchant. Launched 2016-11-23, 148 reviews, 4.9 stars, 92% five-star, with a Built for Shopify badge.

SOptim works at a different point in the funnel. It refuses high-risk sessions at checkout, before the order exists, using a Shopify Validation Function fed by behavior scoring, fingerprinting, and a cross-shop Defense Network. There's no chargeback guarantee, but there's no per-order fee either, and the refused checkouts never become orders — they don't enter your fulfillment pipeline, your refund stats, or your dispute count.

These are complementary models. Wyllo asks: "this order has been placed, should we let it through?" SOptim asks: "this checkout is in progress, should we let it complete?" Stores with high average order values and frequent chargebacks tend to lean toward Wyllo. Stores with high session volume and card-testing pressure tend to lean toward SOptim. Stores with both problems sometimes run both.

The architectural difference

This is the part that decides which app fits which store, so it gets its own section.

Wyllo is a post-checkout service. When an order is placed, Shopify Flow (or a direct Wyllo webhook) fires, the order payload is sent to Wyllo's risk engine, ML scoring runs, and on borderline cases a human reviewer looks at the order. Decisions return as approve, decline, or manual-hold. The merchant's fulfillment workflow waits on that decision, then either ships or cancels. The chargeback guarantee applies to approved orders only — if Wyllo says "ship it" and the order produces a fraud chargeback, Wyllo pays. That's a real financial guarantee and it's the main reason high-AOV merchants choose Wyllo.

SOptim is a during-checkout service. The theme app extension collects behavior signals as the buyer browses and checks out. At the moment Shopify evaluates the cart for checkout, SOptim's Validation Function runs, reads the session's risk score plus fingerprint reputation, and either allows the checkout or refuses it with a message. A refused checkout never becomes an order, never enters the payment processor, never creates an authorisation that could later be disputed. That's the trade-off — no chargeback guarantee, but also fewer chargebacks to guarantee against, and no per-order screening fee.

Both apps use Shopify Functions, but at different hook points and for different decisions. They are not redundant by default.

The feature table

Side-by-side, based on each app's public App Store listing as of 2026-05-22.

Feature	NoFraud / Wyllo	SOptim
Decision point	After order is placed	During checkout (before order)
ML-powered risk scoring	Yes	Yes (behavior + fingerprint + Defense Network)
Human manual review on borderline orders	Yes	—
Chargeback guarantee on approved orders	Yes (paid tier)	—
Server-side checkout refuse (Validation Function)	—	Yes
Behavior scoring (mouse / keys / paste / scroll)	—	Yes
Heavy fingerprint hash (canvas + audio + WebGL)	—	Yes
Cross-shop reputation network	—	Yes (Defense Network)
Country / IP block list	—	Yes
VPN / proxy / TOR detection	—	Yes (Grow+)
Identity verification (3DS, SMS)	Yes	—
Return-fraud / policy-abuse detection	Yes	—
Reseller / bot detection	Yes	Yes
Order-level fraud rules (BIN, disposable email, velocity)	Custom rules engine	Yes (6 templates, Free)
Dispute evidence packets (signed PDF + JSON)	—	Scale+
Lighthouse store audits (LCP / CLS / INP)	—	Yes (Free, unlimited)
Built for Shopify badge	Yes	Not yet
Pricing model	Free to install + per-order fee on paid tier	Flat monthly tiers, no per-order fee
Free tier	First 100 orders/mo free (after 14-day trial)	50 bot blocks/mo + unlimited audits

Sources: Wyllo (NoFraud) App Store listing (apps.shopify.com/nofraud-chargeback-prevention-and-protection, 148 reviews, 4.9 stars, Built for Shopify, as of 2026-05-22) and SOptim's pricing page. Dashes mean the feature isn't documented on the public listing — not that it's impossible.

The chargeback-guarantee question

This is the feature that defines the Wyllo product, and it deserves a clear-eyed look. The model is straightforward: Wyllo runs ML and human review on incoming orders. For orders Wyllo approves on the paid tier, if a fraud chargeback later hits that order, Wyllo reimburses the merchant for the chargeback amount. The economics depend on Wyllo's approval rate being high enough that the per-order fee is less than the avoided chargeback losses.

For a $300 AOV jewelry store, the math often works. A single $300 fraud chargeback that Wyllo would have caught — and is contractually obligated to refund — pays for many months of screening. For a $25 AOV apparel store doing high session volume, the math can be tighter, because the per-order fee scales linearly while chargeback amounts are bounded by AOV.

SOptim doesn't offer a chargeback guarantee. Our model is different: we try to refuse the checkout before the order is placed, so the chargeback never has a chance to happen. We can't promise zero chargebacks — no app can, because some fraud profiles look identical to legitimate buyers during checkout and only reveal themselves at dispute time. What we can do is reduce the flow of fraudulent orders into your funnel, and ship dispute evidence packets (on the Scale tier) that help you win the chargebacks that do happen.

If chargeback losses are your largest fraud cost, and you operate at a price point where Wyllo's per-order fee is comfortably below your AOV chargeback rate, Wyllo's guarantee is a strong proposition. If checkout-level abuse is your problem — card-testing, scraping, repeated failed authorisations clogging your payment processor — SOptim handles that class better, and at flat monthly pricing.

Where Wyllo is the better pick

Three real cases where we'd point you at Wyllo (NoFraud) without hesitation.

• High-AOV stores with regular fraud chargebacks. Jewelry, electronics, luxury apparel, designer goods. The per-order fee economics favor Wyllo here, and the chargeback guarantee converts an unpredictable cost (disputes) into a fixed line item (screening fees).
• Manual review depth. Wyllo's human review team looks at borderline orders and applies judgement that no purely automated system can match. If your store has unusual order shapes — high-value international gifts, B2B drop-shipping with company billing addresses that don't match shipping — manual review catches cases ML alone misses.
• Return-fraud and policy abuse. Wyllo's product covers the post-purchase abuse side of fraud — wardrobing returns, refund abuse, repeat returners on the same payment instrument. SOptim doesn't operate in this category.

Where SOptim is the better pick

Symmetrically, the cases where SOptim handles problems Wyllo's model isn't designed for.

• Card-testing and authorisation abuse. Bots trying stolen card numbers against your checkout, producing dozens or hundreds of failed authorisations per hour. These never become orders Wyllo can screen — they're failed checkouts. SOptim's Validation Function refuses these sessions before the card is even attempted, which means your gateway stops seeing the requests, your auth fees drop, and your Shopify abandoned-checkout list stays clean.
• Scraping and reconnaissance traffic. Competitor price-monitors, content scrapers, vulnerability scanners hitting your storefront. Wyllo only sees orders, so it can't act here. SOptim's behavior signals and fingerprint reputation network catch these sessions on the storefront pages.
• Multi-shop merchants. If you operate three or five Shopify stores, the Defense Network shares blocking signals across them — a fingerprint blocked on Shop A protects Shops B through E for the next 24 hours automatically. Wyllo's screening is per-store and doesn't share cross-store signals by default.
• Flat-fee pricing predictability. If your order volume is variable and you don't want a screening fee that scales linearly with orders, SOptim's flat monthly tiers are easier to budget. The $19/mo Grow tier covers 500 blocks; the $49/mo Scale tier covers unlimited blocks. No per-order math.
• Built-in site audits. SOptim's audit module is unlimited on the Free tier — Lighthouse Performance, Accessibility, SEO, Best Practices, with PDF download. The behavior signal collection and the audit module share one theme app extension. Wyllo doesn't ship audits — different product.

Pricing comparison

Snapshot taken 2026-05-22 from each app's public pricing surface. Wyllo's per-order pricing is plan-dependent; we've left the rate blank where it isn't published on the App Store listing itself.

Tier	NoFraud / Wyllo	SOptim
Free / Starter	Free to install. 14-day trial then first 100 screened orders/mo free	50 bot blocks/mo + unlimited audits + 6 order-rule templates + 24h RUM
Entry paid	Per-order screening fee on orders above the free monthly cap	$19/mo Grow — 500 blocks/mo, VPN+iCloud Relay, 5 custom rules, 30d RUM
Mid paid	Per-order fee + chargeback guarantee add-on	$49/mo Scale — unlimited blocks, 50 custom rules, full Defense Network, dispute evidence
Top paid	Custom enterprise pricing	$99/mo Max — 3 shops, ML auto-tuning, revenue impact, Care tier (2h/mo dev), 4h founder SLA

The two pricing models reflect the two decision points. Wyllo charges per screened order because each screening consumes ML inference and, on borderline cases, human review time. SOptim charges a flat monthly fee because each blocked session consumes a small amount of compute and the Defense Network lookup, both of which are roughly constant per session.

Running both apps

It's allowed and we've seen it work. The configuration we'd suggest is:

• Let SOptim refuse obvious bot and card-tester sessions at checkout. The Validation Function rejects sessions with risk score below your threshold (default 30 on Grow, configurable on Scale). These never become orders, so Wyllo never sees them, and you never pay a per-order screening fee on them.
• Let Wyllo screen the orders that survive SOptim's checkout block. By definition these passed our behavior and fingerprint checks, but they might still be human-driven fraud with stolen credentials — exactly the case Wyllo's ML plus human review handles best.
• If you're on Wyllo's chargeback guarantee tier, the guarantee still applies to orders Wyllo approves. SOptim's checkout block doesn't reduce the guarantee — it reduces the volume of orders Wyllo screens, which usually reduces your Wyllo bill.

Both apps register Shopify Functions, both are well-behaved, and we haven't seen merchants report conflicts running them together. If you do, we'd want to hear about it — drop us a note at /contact.

The decision in one sentence

If your fraud cost shows up as chargebacks on shipped orders, Wyllo's screening plus guarantee is built for that. If your fraud cost shows up as card-testing volume, scraping pressure, or abandoned-checkout junk on your dashboard, SOptim's checkout-time block is built for that. Stores with both problems frequently run both apps.

Footnotes

• Wyllo (formerly NoFraud) review count, star rating, and Built for Shopify badge status taken from apps.shopify.com/nofraud-chargeback-prevention-and-protection on 2026-05-22.
• The app launched as NoFraud on 2016-11-23 and rebranded to Wyllo in late 2025 under vendor Wyllo, LLC. The Shopify App Store URL still contains the nofraud handle for SEO continuity.
• Chargeback guarantee terms are set by Wyllo and apply to approved orders only; merchants should read the current terms on Wyllo's pricing page before committing.
• SOptim is independent and is not affiliated with Wyllo, NoFraud, or its parent company. We have not used Wyllo in production; everything stated here is verifiable on the public App Store listing.
• We will update this page if either app ships meaningful changes. Last review: 2026-05-22.