Service 05

Services · Marketing Attribution

Measurement is only useful if it arrives in time.

Marketing attribution is the practice of connecting a business outcome, a sale, a stream, a ticket purchase, a sign-up, back to the specific marketing activity that produced it. WVE runs it through Centryl, its own platform, so performance data and the decisions made against it sit in the same system: a single command center showing every piece of content, every dollar spent, and every performance metric across every platform while the campaign is running, not in a report that arrived three days later.

Start a campaign See the results

The short answer

What attribution actually does.

The hard part is not collecting the data. It is deciding which activity gets credit, and being honest about the cases where that connection cannot be made cleanly.

Conventional campaign reporting arrives after the money is spent. That makes it a post-mortem rather than a control system. If a campaign runs for six weeks and the report lands in week seven, every decision that mattered was made without data. Attribution, done properly, is a live instrument, not a document produced when there is nothing left to change.

The problem most teams actually face is not that they have no data. It is that their data cannot answer the question that matters: which activity, on which platform, against which audience, produced which result? Without that specificity, spend is held flat rather than reallocated, and the creatives that are not converting keep running until the flight ends.

Centryl exists so that performance data and buying decisions are not in separate places. A campaign that generates signal in the morning is optimized against that signal before the afternoon spend goes out. That is the operational difference.

Where attribution genuinely breaks down

Rigour about methodology is the product here, not a disclaimer. Attribution has real limits, and a page about measurement that does not name them is itself a measurement problem.

Multi-touch journeys, a user who saw a clip, heard a track on a playlist, then searched and streamed, cannot be attributed cleanly to any single touchpoint. Platform walled gardens do not share user-level data across each other. A sale that occurs days after the last tracked impression is not recoverable from click data alone. These are structural limits of the discipline, not gaps in any particular tool.

WVE reports what is measurable directly, labels what is inferred, and does not present correlation as causation. Where a connection cannot be established, the report says so.

Ideal client

Who this is for.

Attribution is useful when there is a real decision it can change. It is not useful as a reporting formality.

01

Brands with meaningful paid budgets

The higher the spend, the more expensive it is to carry underperforming creatives or placements for a week. Daily optimization compounds over the length of a flight.

02

Teams that cannot distinguish between channels

Blended ROAS is a number that can look healthy while one channel subsidises another. Channel-level attribution is what separates the two.

03

Labels & artist teams

Streaming uplift, playlist adds, and presave conversion all happen in short windows. Understanding which campaign activities drove which metric, even partially, informs how the next release is built.

04

Live events & festivals

Ticket sales have a clean conversion event and a short window. When link tracking is in the distribution path, attribution is among the more tractable problems in marketing.

05

Direct-to-consumer businesses

Site traffic that does not convert is a measurement problem before it is a media problem. Knowing where users drop off changes which problem gets fixed first.

06

Not a fit

Campaigns with no tracked conversion event, or where the product or pricing is the limiting factor rather than the media. Attribution cannot fix a destination that is not working. We will say so before taking the work.

Process

How we run it.

Five stages. The first is usually the one that reveals how much of a team's current reporting is architectural rather than informational.

1. Measurement architecture, before the campaign launches

Before any creative goes live, WVE establishes what is trackable, what is inferable, and what cannot be attributed. Link tracking, UTM taxonomy, conversion events, and audience pixels are set up consistently across every platform in the campaign. This is the work that determines the quality of everything that follows, running it after the campaign starts means the first two weeks are unrecoverable.

2. Real-time tracking across platforms

Campaign performance is pulled into Centryl across platforms in a single view: reach, impressions, click-through, conversion by placement, and spend by channel. The operative word is single, when these sit in separate platform dashboards, comparing them requires manual export, normalisation, and time, all of which defeats the purpose.

3. Daily optimization against live signal

Spend is moved between creatives, audiences and placements daily while the campaign is live. A creative that is not converting stops consuming budget the same day that becomes clear. A placement that is outperforming gets more weight before the next morning's spend goes out. The plan agreed at kickoff is a starting allocation, not a commitment.

4. On-site behaviour and destination analysis

Traffic that arrives and leaves is a measurement problem before it is a media problem. Monitoring how users actually behave on site, where they stop, what they ignore, where they abandon, identifies whether a weak conversion rate is a media issue or a destination issue. These are different problems with different solutions.

5. Post-campaign attribution report with stated limits

The final report distinguishes between what was directly attributed, what was correlated, and what could not be connected. Every figure carries its method. Where a causal claim cannot be supported, the report says so and explains why. This is the document that makes the next campaign cheaper to plan.

What you receive

Live Centryl dashboard

Every channel, creative, and placement in one view, updating continuously while the campaign runs. Spend, reach, conversion, and efficiency by line item, not a blended total.

Channel-level attribution breakdown

Which platforms and placements drove which outcomes, separated by directly attributed versus inferred contribution. The blended figure is also there; it is just not the only figure.

On-site behaviour analysis

Where campaign traffic lands, what it does, and where it drops. Presented as a ranked list of the changes that would have the highest impact on conversion, not as a dashboard requiring interpretation.

Methodology statement

A plain-language explanation of what was tracked, how each figure was produced, what was inferred, and what could not be attributed. This is the document the next agency will ask for.

Named case studies

Campaigns you can check.

Named clients, stated outcomes. Each headline result is one WVE can substantiate. Where a figure comes from blended reporting across channels rather than from a single clean attribution model, this page says so.

Moncler #WarmerTogether · brand

A trending campaign where the attribution model was velocity rather than conversion: the goal was a specific platform threshold within a specific window. Reaching No.1 trending in the United States is a binary outcome, one of the few on this page that does not require a measurement window caveat.

No.1 trending, United States

Scale & methodology

The numbers, with their working.

A figure without its window and its method is a claim, not evidence. These are the most carefully stated metrics on this site because this page is about measurement and consistency is not optional.

$4B+: Client sales attributed Cumulative since 2019, across all client campaigns. Basis is blended spend-to-revenue as reported per campaign. Attribution model and window not yet published, read as directional, not as an audited total.
18.6×: Average paid-media ROAS Average across paid-media campaigns since 2019, platform-reported. Attribution model and measurement window not yet published. Not normalised across campaign types or sectors.
25B+: Accounts reached Cumulative unique accounts exposed to a client campaign since 2019, as reported by each platform. Not de-duplicated across platforms, a user reached on two platforms counts twice.
330+: Artists on platform Distinct artist accounts active on Centryl since 2019. This is a client count, not a campaign count, one artist may have run multiple campaigns.

Read this before quoting any figure above. All four are cumulative totals from WVE's capabilities deck. The ROAS and client-sales figures are the most cited on this site and the most methodologically exposed: neither yet carries a published attribution model, a measurement window, or a breakdown by channel or campaign type. WVE is aware of this gap, naming it here is deliberate. For these metrics to be audit-grade, WVE would need to publish: (1) the attribution model used per campaign type, (2) the measurement window applied, (3) whether figures are last-click, blended, or correlation-based, and (4) the range across campaigns rather than the average alone. Until that is done, these figures should be read as directional indicators.

FAQ

Questions we actually get.

These mirror the phrasings people and engines use, and each answer matches the FAQPage schema on this page word for word.

What is marketing attribution?

Marketing attribution is the practice of connecting a business outcome, a sale, a stream, a ticket purchase, a sign-up, back to the specific marketing activity that produced it.

The hard part is not collecting the data; it is deciding which activity gets credit when a customer touched multiple channels before converting, and being honest about the cases where that connection cannot be made cleanly.

How do you attribute a stream or a ticket sale to a clip?

Imperfectly, and this page will not pretend otherwise. For a stream, the causal chain is: a user sees a clip on a social platform, follows through to a streaming service, and streams the track. The first step is measurable at the platform level via link tracking and UTM parameters. The second step depends on whether the streaming platform passes back conversion data, which varies by platform and deal. The third step, isolating the clip's contribution from every other thing the listener heard that week, cannot be attributed cleanly.

What WVE tracks is the correlation between clip distribution volume and streaming uplift in the same window, across the same markets, and reports that as an observed relationship rather than a proven cause. For ticket sales the chain is shorter and cleaner when a tracked link is in the distribution path; longer and noisier when a user converts days later from organic recall.

What is Centryl?

Centryl is WVE's own measurement and optimization platform. It exists so that campaign performance data and the decisions made against that data live in the same system, which is what makes optimising a campaign while it is still running practical rather than theoretical.

A single command center shows every piece of content, every dollar spent, and every performance metric across every platform while the campaign is running, not a report that arrived three days later.

What does WVE's 18.6x ROAS figure mean, and what does it not mean?

18.6x is the average paid-media return on ad spend reported across WVE's campaigns. It is a real figure from WVE's capabilities deck. What it does not yet carry is a published attribution model, a stated measurement window, or a breakdown by channel or campaign type.

On a page about attribution, citing it without those qualifications would be the exact problem this service exists to solve. WVE is aware of this and is working towards publishing the methodology. Until that is done, the figure should be read as a directional indicator, not as a benchmark against which other agencies can be compared.

Can WVE work with our existing analytics stack?

Centryl is designed to sit alongside rather than replace existing client-side analytics. The specific integrations supported, and whether Centryl supplements or supersedes tools like GA4, depend on the campaign setup and are confirmed at onboarding.

If you are running a particular stack, name it in the brief and WVE will confirm the connection before the work starts.

How quickly does optimization happen once a campaign is live?

Continuously. Spend is moved between creatives, audiences and placements on a daily basis while a campaign is live, rather than being reviewed at the end of a flight.

The practical result is that a creative that is not converting stops consuming budget the same day that becomes clear, rather than a week later when the report arrives.