Connecting Magento to Roku Ads

Architecture

How Magento and Roku actually connect

Two systems that know nothing about each other. Everything below exists to give them one shared reference.

1. Step 1

   Magento holds the truth

   You own the database, so the truth is as good as the query. Pin down which order statuses count as revenue — processing, complete, or both — because that choice moves the number substantially.

2. Step 2

   Roku Ads needs a click to match

   A pixel or server event on your side that Roku can match against, plus the hashed customer information the match depends on. Without that, the channel is measurable only by holdout.

3. Step 3

   The identifier is the bridge

   Household-level identity, not a click, captured at the visit and carried into the Magento order. Without it there is nothing to join on and no tool fixes that.

4. Step 4

   The event source decides reliability

   Fire server-side from the observer rather than from the success page template, which is cached and may not execute anything.

5. Step 5

   Deduplication at the destination

   The deduplication problem is across platforms, not within this one. A household exposed to a TV ad and later reached by search and social produces a conversion all three will claim.

Diagnose it honestly

Working vs. broken, side by side

Both states look identical from the Roku dashboard, which is why this runs for months before anyone notices.

Diagnostic sequence

Diagnostic order

Each step rules out a class of cause. Run them in order and the last one usually has a single explanation standing.

1. 1

   Count orders carrying household-level identity, not a click

   The ceiling on everything else. Do this before debugging anything downstream.

2. 2

   Follow one real conversion end to end

   From the ad click through Magento and into Roku.

3. 3

   Place a test order as an anonymous visitor through cache

   Not as an admin, which bypasses it. If the order data is missing on the cached path, caching is your cause.

4. 4

   Ask whether a geo or audience holdout exists

   If it does not, the reported figure is exposure correlated with purchase, which is not the same as a campaign that caused one.

5. 5

   Read the match rate, not the success message

   Roku reports matched conversions and match rates. The match rate tells you how much of your audience it could resolve at all — and a geo holdout tells you whether the campaign moved anything.

Source-of-truth validation

How it gets proven

Against Magento's own records, not against either platform's dashboard.

• Household-level identity, not a click confirmed present on a live order.
• A real conversion followed from click through Magento to Roku.
• Incremental lift measured against a holdout, with platform-reported conversions shown alongside rather than instead.
• Order data confirmed populated on an order placed through a fully warmed cache.
• Reconciled over a fully closed window, matched per record rather than on totals.
• Re-checked 30 days later.

Straight talk

When you don't need this

Worth reading before paying anyone, including us.

• If Roku already reports conversions that reconcile against Magento's records within a few percent, this join is working and there is nothing to buy.
• Measure identifier coverage yourself first — pull your last hundred orders and count how many carry household-level identity, not a click. Ten minutes, free, and it tells you whether any of this is worth doing.
• If your Magento build has heavy customisation by another team, scope their availability first — the build is usually blocked there.
• If you cannot run a holdout, keep CTV budget small and report it as unproven. That is a more useful position than a matched number nobody can check.