# ReferralCandy vs Viral Loops: which do AI models recommend for referrals, September 2026

GTM AI Recommendation Index, September 2026 Edition, Referral and partner programs. Zero of twelve models named ReferralCandy first on the direct prompt; zero named Viral Loops. Page: https://gtm-ai-index.com/partner/referral-and-partner-programs/referralcandy-vs-viral-loops/

| | First-choice share | Rank | Negative rate | Labels | Models naming it |
|---|---|---|---|---|---|
| ReferralCandy | 10% | #3 of 11 | 10% | 31 | 12 of 12 |
| Viral Loops | 4% | #5 of 11 | 20% | 20 | 10 of 12 |

## The direct prompt, model by model

- Claude Haiku 4.5: neither named (first choices: Cello) (alternatives: Friendbuy, Influitive, Partnero, Referral Rock)
- GPT-5.4 mini: neither named (first choices: Referral Rock) (alternatives: Friendbuy, Referral Factory)
- Gemini 3.5 Flash: neither named (first choices: Referral Rock) (alternatives: GrowSurf, PartnerStack, impact.com Advocate)
- Perplexity Sonar: neither named (first choices: Ambassador) (alternatives: CustomerGauge, GrowSurf, Referral Rock)
- Grok 4.1 Fast: neither named (first choices: Referral Rock) (alternatives: Ambassador, GrowSurf, Influitive)
- Mistral Small: neither named (first choices: Referral Rock) (alternatives: Ambassador, Cello, GrowSurf)
- DeepSeek V4 Flash: neither named (first choices: Referral Rock) (alternatives: Ambassador, GrowSurf)
- Llama 4 Maverick: neither named
- Qwen 3.7 Flash: neither named (first choices: Cello) (alternatives: Friendbuy, GrowSurf, Referral Rock)
- Kimi K2: neither named (first choices: Referral Rock) (alternatives: Ambassador, GrowSurf, Influitive)
- GLM 4.7 FlashX: neither named (first choices: Ambassador, PartnerStack) (alternatives: GrowSurf, Referral Rock, impact.com)
- MiniMax M2.5: neither named (first choices: Referral Rock) (alternatives: Mention Me, impact.com)

## What the models said about ReferralCandy

- "What to avoid for mid-market B2B" (DeepSeek V4 Flash, direct prompt, hard negative)
- "it historically charges a 1.5% to 10.5% "success fee" on e-commerce sales generated through referrals... it becomes incredibly expensive as your brand grows." (Gemini 3.5 Flash, negative prompt, soft negative)
- "*Cautious approach to pricing model* While generally well-reviewed (4.9/5 on Shopify), be aware of: Success fees" (DeepSeek V4 Flash, negative prompt, soft negative)
- "Most budget-constrained companies find ReferralCandy or ReferralHero provide the best balance of features versus cost" (Qwen 3.7 Flash, budget prompt, first choice)
- "If you can spend ~$40-50/month: ReferralCandy offers the best balance of features and ease-of-use for e-commerce" (Kimi K2, budget prompt, first choice)
- "For most small-to-medium businesses, ReferralCandy or Viral Loops offer the best balance of features and ease of use" (MiniMax M2.5, comparative prompt, first choice)

## What the models said about Viral Loops

- "Multiple reports of billing issues ... 1-star reviews highlight "deceptive practices" and "waste of money."[" (Grok 4.1 Fast, negative prompt, hard negative)
- "aimed at B2C viral loops and pre-launch waitlists, not ongoing B2B referral operations" (DeepSeek V4 Flash, direct prompt, hard negative)
- "*Caution for cash reward programs* — No automated fraud detection" (DeepSeek V4 Flash, negative prompt, soft negative)
- "Start with Viral Loops - it offers the best combination of affordability ($35/month), comprehensive features, and a solid free trial period." (GLM 4.7 FlashX, budget prompt, first choice)
- "ReferralCandy or Viral Loops offer the best balance of features and ease of use" (MiniMax M2.5, comparative prompt, first choice)
- "Viral Loops (≈$35/month) is the best all-around value" (DeepSeek V4 Flash, budget prompt, first choice)

Share is the count of first choices across the direct, paraphrase, budget and scale prompts over all twelve models, for a mid-market B2B company; rank is within the category. Comparisons are drawn for the top eight products in each category. Published under CC BY 4.0; the output is the models' output, and nothing here is a recommendation by the index.