FMCG D2C: the analytics rebuild that pays in 6 months
Klaviyo + Mixpanel + Segment + Shopify Plus add up fast. The rebuild ROI is now under 12 months.
If you’re running a D2C FMCG brand at $20M–$100M GMV, your analytics + retention stack probably looks like this: Klaviyo for email/SMS, Mixpanel or Amplitude for product analytics, Segment for event routing, Shopify Plus for the storefront, plus three or four “best of breed” tools for attribution, customer reviews, loyalty, and post-purchase. Combined annual spend is $300K–$900K. None of the tools individually feel expensive; the combined bill is the problem.
The rebuild ROI on this category has tightened in 2024–2026. Most rebuilds we’ve shipped pay back in 6–9 months, with Y2+ reclaim in the $200K–$600K range.
Why the math has tightened
Three vectors:
- Per-contact pricing on email/SMS (Klaviyo, Attentive, Postscript) compounds fast above 100K active subscribers. The marginal cost of an email send is essentially zero on the underlying provider (SES, Postmark, Twilio). The licence margin is doing the work.
- Per-event pricing on product analytics (Mixpanel, Amplitude) compounds as event volumes grow. PostHog and self-hosted alternatives are operationally cheap once installed.
- Segment pricing has scaled aggressively at enterprise tiers. Open-source CDPs (Rudderstack, Jitsu) cover the same routing function at meaningful cost reduction.
A typical Klaviyo + Mixpanel + Segment combo at $80M GMV runs $35K–$60K/month. A rebuild covering the same functions runs $4K–$8K/month in infra, with a one-time build cost of $90K–$150K.
The reference rebuild
Here is the architecture we ship most often for D2C FMCG brands in this category:
| Function | Replacement | Notes |
|---|---|---|
| Email + SMS sends | AWS SES + Twilio (or Resend + Twilio) | Direct providers, no per-contact margin |
| Email/SMS authoring + scheduling | Custom UI on top of SES/Twilio | 4-week build, uses internal Postmark-style admin |
| Product analytics | PostHog (self-hosted on AWS) | Replaces Mixpanel; team-owned data |
| Event routing | Rudderstack open-source | Replaces Segment; same SDK contract |
| Storefront | Shopify Plus (kept) | Rebuilding storefront usually doesn’t pencil |
| Reviews, loyalty | Yotpo / Stamped (kept or replaced individually) | Case-by-case |
| Attribution | Custom multi-touch model on PostHog events | Replaces Northbeam / Triple Whale at scale |
We do not replace Shopify Plus. The storefront is the one component where the buy economics still win — Shopify’s cost-to-serve is genuine, and the operational cost of a custom storefront is high.
What changes operationally
A few things that brand operators should expect from the rebuild:
- The marketing team retains creative authority. Email templates, SMS copy, segmentation logic — all live in the custom UI, owned by the marketing team. The rebuild shifts the platform, not the workflow.
- First-party data ownership becomes meaningful. PostHog stores events in your AWS account. Customer profiles live in your database. The exit cost from the rebuild is essentially zero, where the exit cost from Klaviyo is real (data export, contact-list rebuild).
- Deliverability requires attention. Klaviyo/Attentive maintain sender reputation across their entire customer base. On AWS SES you maintain your own. This is not difficult but does require ~2 hours/month of operational discipline.
- Reporting is custom, not templated. Klaviyo’s standard reports go away. Custom dashboards in Metabase or Grafana replace them. Most teams find this an improvement once they’ve adjusted; some teams prefer the templated experience.
A representative engagement
A recent ASX-aligned D2C brand, ~$60M GMV, was paying $42K/month combined Klaviyo + Mixpanel + Segment + Northbeam. We rebuilt the stack in 12 weeks at $128K fixed price. New monthly run cost: $5,800. Y1 reclaim: $307K. Y2+ reclaim: $435K/yr.
The brand’s CFO had flagged the marketing-tech bill in three consecutive quarterly reviews. The CMO had defended it on grounds of vendor maturity. The rebuild conversation only became serious when the engineering team realised the rebuild was a fixed-price 12-week engagement, not a 9-month internal project.
Where the rebuild doesn’t fit
Three cases where we’ve talked brands out of the rebuild:
- <$10M GMV. The fixed cost of build doesn’t recover fast enough. Stay on the SaaS stack.
- No in-house engineering capacity, even via vendor. The rebuilt stack requires ongoing operational ownership. If the brand’s engineering function is one part-time contractor, SaaS is the safer answer.
- Heavy reliance on integrations the SaaS does well. If the brand’s value chain depends on a Klaviyo integration that doesn’t exist in the open-source stack, the integration cost erases the savings.
Outside these cases, the math is increasingly hard to argue against.
How to evaluate
We do a free 30-minute analytics-stack review. We pull your current bill (you share in the call), map it against the rebuild architecture, and produce a back-of-envelope payback estimate. If the payback is >18 months we tell you so and walk away. If it’s <12 months, we quote a fixed-price rebuild.
Read more: /upstream/klaviyo-alternative · /upstream/mixpanel-alternative · /sectors/fmcg · /sectors/ecommerce
Run the matching free calculator
Each one runs in 3 minutes and emails you an 8-page memo.