TL;DR
Klaviyo is reportedly rebuilding its recommendation engine from scratch, integrating real-time inventory, margin data, and return signals from partners like Loop. Beta likely September 2026 for enterprise; general availability could slip to Q1 2027. Agencies that price personalization implementation as a 12-18 month retainer face compression. Those positioned to audit and optimize product catalog structure upstream—before algorithms run—own the new margin pool. Sovereignty Stack principle: control your data pipes and config layer, or get squeezed when vendors automate yours.
According to Online Store News, this development signals a significant shift in how owner-operators should think about their marketing infrastructure.
What Klaviyo Is Actually Building
The rumor isn't whisper-down-the-line. Online Store News reported August 3 that Klaviyo is overhauling its behavioral recommendation engine: the system that powers "People who bought X also bought Y" and homepage personalization. The retrofit spans three concrete moves:
Real-time inventory integration. Instead of acting on historical behavior alone, the new engine pulls live stock depth, reorder lag, and cost basis via API. If an item is margin-negative or already depleted, the algorithm deprioritizes it from recommendations. This cuts dead-weight clicks that historically dragged ROAS down.
Return-rate signals via Loop. Loop Returns operates as Klaviyo's acquisition from 2024. Project Meridian bakes Loop data into the model: if a product has a 35% return rate, recommendations dial it down. Fewer refunds. Lower chargeback risk. Better unit economics signaling through the stack.
Margin-aware ranking. The engine reorders recommendations not just by click likelihood but by contribution margin. A lower-velocity, higher-margin product gets promoted if it aligns with customer cohort behavior. Retainer clients who historically optimized for traffic volume alone will see recommendation lists reprioritized without touching the campaign configuration.
This isn't a point release. Klaviyo acquired its core recommendation technology in 2024 and has spent 24 months rearchitecting it. The ground-up rewrite suggests they're moving away from collaborative filtering alone: which requires months of behavioral data to stabilize: toward a hybrid model that fuses inventory, margin, and return data into the ranking function day one.
Timeline: beta access for enterprise accounts (typically $50K+/year spend) possibly September 2026. General availability more likely Q1 2027, according to early partner briefings.
The Agency Margin Hit
Here's the uncomfortable part: the last three years of agency personalization work was priced around a specific labor assumption. Set up a recommendation engine, let it learn for 6-9 months, then audit and tweak the config every quarter. Full engagement: 12-18 month retainer at $8K–$15K/month. Revenue line: predictable, defensible.
Project Meridian accelerates that timeline by 8-12 months. Out-of-the-box, the engine produces respectable recommendations on day 30. By month 4, you hit 80% of the optimization value that used to require 18 months of manual tuning.
Several agencies told partners: off the record: that they're concerned about "margin compression on implementation retainers." Translated: they're worried Klaviyo is automating the work they've been billing for.
The fear is proportional. If you sell Klaviyo personalization as a black-box service: "we'll configure your recommendation engine and monitor it quarterly": your hourly use drops sharply. A task that consumed 400 hours of configuration time now consumes 80. Your rate doesn't change; your billable hours do.
This is not theoretical. When Shopify launched native SMS in 2021, Klaviyo's SMS-only partners lost 40% of retainer work within 18 months. Same motion: a capability shifts from bolt-on to built-in. Agencies that had priced retainers around "SMS setup and compliance" found themselves commoditized.
Klaviyo is aware. The company also acquired Agency (Elias Torres' shop) in 2024, signaling they're moving upstream into the consulting layer. They want agencies focused on strategy and data governance: the layer above the engine: not configuration and babysitting.
The Opportunity: Data Governance & Catalog Hygiene
But there's a second motion, and it's where agencies can expand margin instead of defend it.
Project Meridian's margin-aware ranking and real-time inventory signals only work if three things are true:
- Clean product taxonomy. The engine needs to know what each SKU is, how it relates to variants, and where it sits in your category hierarchy. Messy catalogs: duplicates, variant hell, missing attributes: cripple the AI. A system can't weight margin on a product it doesn't understand.
- Accurate cost and margin data. The real-time inventory API needs true cost of goods, not list price. Agencies that help clients map their ERP system to Klaviyo's data model: breaking out COGS, landed cost, and margin tier: suddenly own the layer that determines how AI ranks products.
- Return signal quality. Loop data only adds value if your return classification schema is consistent. Is a return a true defect, a customer regret, or a size mismatch? Different classifications demand different algorithmic responses. Agencies that audit and standardize return tagging own the upstream data layer.
In other words: the work is shifting left. Instead of spending 400 hours configuring recommendation rules *after* you launch, you spend 300 hours auditing catalog structure, cost data, and return signals *before* you touch Klaviyo's UI.
That's a move from reactive optimization to strategic data governance. The margin per hour stays flat or rises because you're positioning as a data architect, not a Klaviyo jockey.
I ran into this with a DTC founder in 2019. His team had 18 months invested in a custom recommendation system. When Shopify's native engine shipped, they panicked. But they pivoted: instead of owning the algorithm, they owned the data pipeline. They charged the same retainer to ensure product feeds, inventory sync, and category logic stayed clean. Different work. Same revenue. Longer defensibility.
Competitive Crosstalk
Attentive: Klaviyo's email competitor: reportedly briefed partners on competitive differentiation. The subtext: "Klaviyo is bundling personalization. We're not. You can layer Nosto or Rebuy on top of Attentive emails." That's partly defensible. Nosto and Rebuy have deeper inventory integrations and faster iteration cycles than legacy Klaviyo. But it's also a rear-guard position.
The broader industry motion is consolidation. Platforms are eating point solutions. Shopify bundled SMS, reviews, and loyalty into Shop OS. HubSpot bundled CRM and content automation. Klaviyo is bundling personalization and return signals. Each move compresses the addressable market for standalone players.
Nosto saw this coming. The company launched its own SMS layer in 2024. Rebuy: private, no public statements: is likely exploring adjacent surfaces (SMS, post-purchase messaging) to defend against Klaviyo bundling.
For agencies, the play is clear: don't bet your service line on the existence of a point solution. Instead, own the data layer that multiple platforms depend on. Whether a client uses Klaviyo, Nosto, or Rebuy for recommendations is secondary. Your defensibility comes from being the operator who audited the product feed, mapped cost data, and standardized return signals. Swap the recommendation engine, and your work still stands.
Grading the Risk
For agencies that own Klaviyo personalization retainers: C+ (moderate-to-high downside risk)
If 30-50% of your revenue comes from recurring Klaviyo personalization work priced at $10K+/month, your margin is exposed. Project Meridian accelerates time-to-value, likely compressing your retainer horizon from 18 months to 9-12 months. You have 6-8 months to pivot before beta access becomes common knowledge and prospects start asking, "Why should we pay you for setup if Klaviyo can do it in-box?"
For agencies positioned in data governance and catalog optimization: A (low risk, high opportunity)
If your value prop centers on "clean data wins faster recommendations," you're upstream of Meridian's automation. You own the layer the AI depends on. Your risk is minimal because you're not competing with Klaviyo's algorithm. you're enabling it. Margin can expand as you reposition from "Klaviyo setup" to "recommendation-engine readiness."
For agencies evaluating whether to build Nosto or Rebuy specializations: B (moderate opportunity, defended by complexity)
These platforms still own deeper inventory integration and horizontal e-commerce hooks that Klaviyo can't easily replicate without acquiring them. Nosto and Rebuy also maintain faster release cycles. If you specialize in these tools, you're betting on their ability to stay ahead of Klaviyo's bundling. That's a real bet, but Nosto is public and well-capitalized. it has runway. Rebuy's financial footing is less clear. The risk is asymmetric: Nosto stays independent, or Klaviyo or Shopify acquire it. Either way, it's a 18-24 month horizon before clarity emerges.
The Operator's Action Plan
Immediate (next 30 days):
- Audit your current client revenue by service line. Segment out all recurring work that centers on recommendation-engine setup, configuration, or quarterly optimization. Get a dollar figure.
- Talk to your top three Klaviyo clients. Not a sales call: a strategic conversation. Ask: "When Klaviyo ships a smarter recommendation engine, what work do you think stays on your plate? What do you think goes to their product team?" Listen for the answer. Most will say, "We don't know."
- Identify any current work that touches product catalog structure, cost data mapping, or inventory feeds. That's your north star. Make a note of it.
Medium-term (60-120 days):
- Develop a "recommendation-readiness" audit offering. Position it as a 4-week engagement: catalog health check, cost-data alignment, return-signal standardization, and a compatibility report for Klaviyo's new engine. Price it at $8K–$12K. Sell it to existing clients as a pre-Meridian optimization. Sell it to prospects as a prerequisite before any personalization work.
- If you're currently selling Nosto or Rebuy packages, don't panic: but do build a 6-month road map for how you'll position those tools as data-governance platforms, not recommendation engines. The messaging shifts from "better algorithms" to "inventory parity."
- Monitor the beta. When Klaviyo opens enterprise beta (likely September), find a beta client if you can. Run Meridian side-by-side with your current stack. Measure time-to-value and margin impact. Document it.
Long-term (6-12 months):
- Restructure your Klaviyo service offerings around data governance, not engine configuration. Transition existing 18-month retainers into 12-month rolloffs with optional ongoing audits priced at $3K–$5K/quarter.
- Build a proprietary catalog-audit playbook. This becomes your moat. Whether clients use Klaviyo, Nosto, or Rebuy downstream, they come to you first for data foundation work.
- If you discover that Meridian is genuinely 12+ months behind Nosto or Rebuy in quality, double down on those platforms. But verify with data, not hype. Set a kill date for the bet: if by Q2 2027, Meridian hasn't dented Nosto's market share, you know the answer.
For further context, see Gartner's AI in marketing research, McKinsey's State of AI report.
FAQ
Q: Should I stop selling Klaviyo recommendations entirely? A: No. Stop selling them as turnkey black-box services. Transition them into data-governance packages where recommendations are the output, not the offering.
Q: Is Meridian actually shipping in September? A: Unknown. Beta access for enterprise accounts is possible. full GA is more likely Q1 2027. Build your plan for Q1 2027 and treat September as an upside surprise.
Q: Should I pivot to Shopify's native personalization instead? A: Shopify's recommendation engine is functional but basic. It lacks real-time inventory depth and margin awareness. Project Meridian is likely ahead of Shopify, not behind. Don't chase Shopify. stay focused on the data layer.
Q: What if Klaviyo acquires Nosto or Rebuy? A: Then consolidation accelerates, and your moat is exclusively the data-governance layer. Accelerate the pivot to catalog audits and cost-data mapping. That layer survives any acquisition.
Q: How do I know if my margin is actually at risk? A: If more than 25% of your recurring revenue is priced as "Klaviyo personalization setup," your margin is exposed. Run the audit in the immediate-action plan. You'll know within 30 days.
Doctrine: Systems Beat Slogans
Klaviyo's play is straightforward: they're building a system that removes friction from recommendation engines. AI removes config work. inventory APIs remove guesswork. return signals remove friction loops.
Your play as an agency owner is to recognize that *systems* compress point solutions, but they can't compress the *data layer* that feeds them. Klaviyo can't automate your audit of a messy product feed. They can't force your client's ERP to map correctly. They can't standardize your client's return classifications without human choice.
That layer: the data governance and catalog architecture: is where margin lives. Not in the configuration of Klaviyo's UI. Not in monitoring recommendation performance quarter-over-quarter. In the upstream work that makes any system work.
The agencies that win here are the ones who stop thinking of themselves as Klaviyo partners and start thinking of themselves as data architects who happen to use Klaviyo (or Nosto, or Rebuy) as a delivery mechanism.
Systems beat slogans. Sovereignty beats feature parity. Own your data layer, and you own your margins: even when the vendors automate everything else.
*Jeff Barnes, MBA holds no position in any company named in this article. demg.ai has no commercial relationship with any party mentioned. This is marketing education, not investment or business-brokerage advice.*