Podcast Prep / Liam Millward / Instant

The hidden retention leak inside every Shopify brand.

A tactical episode about how Instant works alongside Klaviyo: identifying more high-intent shoppers, sending better signals into the existing lifecycle stack, and measuring what is truly incremental without turning the conversation into an anti-Klaviyo pitch.

01 / Thesis

This should not be an AI email marketing episode.

The stronger angle: Instant is a layer on top of Klaviyo and the existing retention stack. The episode should show how the two work together: Klaviyo remains the system of record and messaging engine; Instant helps identify more high-intent shoppers, enrich event signals, and automate the decisions teams do not maintain manually.

01

Eligibility before optimization

Before optimizing subject lines, design, timing, or discounts, ask whether the shopper is even eligible to enter the flow.

02

AI without subject-line slop

AI is only interesting if it makes lifecycle decisions the team was not maintaining manually: product context, timing, offer logic, flow path and intensity.

03

Measurement discipline

The credibility section is incrementality: what is ESP-attributed, what is holdout-tested, and what actually shows up on contribution margin.

02 / The Tension

Make Liam prove the wedge without dunking on Klaviyo.

This is where the episode becomes useful instead of sponsored: respectful to the existing stack, clear on where Instant adds leverage.

Common brand belief
“We already have Klaviyo flows, so the lifecycle stack is covered.”
Liam counter

Klaviyo is the foundation. Instant’s argument is that brands can send Klaviyo more complete identity, intent and product signals, then automate the decisions teams do not maintain manually.

Nik counter-push

How much of this is actually incremental revenue versus attribution dressed up as retention?

03 / Research Notes

What we know.

Use these as prep, not script. Company-reported numbers should be described that way on air.

Product wedge

Instant Audiences identifies opted-in shoppers and sends more high-intent abandonment events into Klaviyo or the ESP. Instant AI can automate campaign/flow decisions around those signals. The point is partnership with the stack, not replacement rhetoric.

Claims to caveat

  • “Incremental revenue” needs definition.
  • Identifiable does not always mean emailable.
  • AI agents are still the beta/future story.
  • 30% email revenue can be healthy or bloated.

Proof points to mention carefully

ThirdLove, Neuro, Kind Patches, FAYT, July and Nakie appear in Instant materials with strong revenue/ROI claims. Treat them as Instant case-study data unless Liam explains holdouts.

04 / Run of Show

A credible, tactical structure.

Start with the spicy metric debate, move into the hidden leak, then force concrete AI and measurement examples.

00:00

Cold open

“What is the biggest lie brands tell themselves about email revenue?” Then: “If a brand says 35% of revenue comes from email, when is that impressive and when is it a red flag?”

03:00

Founder story tied to the problem

Checkout, frictionless commerce, abandoned revenue, raising early, and why Instant moved toward retention and shopper identification.

10:00

The hidden leak: eligibility before optimization

What high-intent shoppers are missed even when Shopify and Klaviyo are already live? What causes the miss?

20:00

Abandonment is not one flow

Collection view, product view, add to cart and checkout start are different intent levels. The message, timing and offer should not be the same.

30:00

AI that is actually useful

Force specificity: copy, product selection, offer, timing, flow path, approval rules and where humans stay in the loop.

40:00

Measurement and incrementality

Define incremental. Compare attributed revenue, holdout-tested lift, modeled revenue and true net-new contribution margin.

52:00

The $5M, $20M, $100M retention audit

Audit three fictional brands: apparel, supplements and beauty. For each, ask what Liam checks first, what flows should exist, what Instant would not fix, and what is a bad fit.

65:00

AI agents, but concrete

What is the difference between an AI agent and a sophisticated flow builder? What campaign did an agent actually plan and execute?

75:00

Closing checklist

Five things every founder should check this week: identifiable traffic, trigger coverage, duplicate sends, margin/offer logic and a holdout plan.

05 / Best Questions

Ask these.

Designed to pull frameworks, not talking points.

Sharp openers

  • What is the most misleading email marketing metric?
  • When is 30% of revenue from email actually bad?
  • What do brands think retention means, and what does it actually mean?
  • If a brand already has Klaviyo, where can Instant add signal, identity or automation on top?

Product clarity

  • In one sentence, how does Instant work with Klaviyo?
  • Are you feeding Klaviyo, enriching events, automating specific flows, or running net-new workflows alongside it?
  • What changes in the account on day one?
  • What does the marketer see inside Klaviyo after Instant is live?

Operator value

  • What are the three flows every $5M+ Shopify brand should have?
  • What flow do brands think is done but almost always underperforms?
  • What data hygiene issues ruin AI retention performance?
  • What is the wrong customer for Instant?
  • Which categories see the biggest lift: apparel, supplements, beauty, furniture, fitness?

Trust and compliance

  • How do you define opted-in?
  • Are there cases where you can identify someone but should not email them?
  • How do unsubscribes and suppression lists sync?
  • What should a legal team ask before turning this on?

Incrementality

  • How do you separate real lift from attribution?
  • What is the cleanest holdout test?
  • What traffic level does a brand need for a useful test?
  • How should brands calculate ROI after discounts, send costs and gross margin?
06 / Segment Ideas

Games that force specificity.

Good sponsored episodes need motion. These bits keep Liam from drifting into vendor language.

Flow or Fiction

Nik names a tactic. Liam says whether it is useful or theater.

  • 15% abandoned cart discount
  • Browse abandonment emails
  • Price drop emails
  • AI-generated copy
  • Send-time optimization
  • “Email drives 40% of revenue”

The Leak Map

Nik gives a funnel stage. Liam names the hidden leak and the fix.

  • Paid click
  • Product view
  • Collection view
  • Add to cart
  • Checkout start
  • Second purchase
Best segment

Stack Upgrade Courtroom

Nik defends the current stack: “I already have Klaviyo, an agency and abandoned cart flows.” Liam has to prove where Instant adds signal or automation on top without making Klaviyo the villain.

Closing checklist

  1. Identifiable traffic percentage
  2. Trigger coverage by intent
  3. Flow overlap and duplicate sends
  4. Offer and margin logic
  5. Holdout or suppression test plan
07 / Pull Quotes

Lines to chase.

These are not scripted lines. They are the soundbites the conversation should earn.

“The biggest retention leak is not bad copy. It is shoppers who never become eligible for the flow.”
“Klaviyo is the foundation. The question is what signal and automation you can layer on top.”
“AI in lifecycle is only interesting if it makes decisions the team was not maintaining manually.”
“A better cart email does not matter if the shopper never enters the cart flow.”
“The job is not to send more emails. The job is to recover more high-intent moments without training customers to wait for discounts.”
08 / Source Links

Research trail.

Useful pre-read. Not all need to be mentioned on air.