What each podcast advertising attribution method means for your numbers

What to take away

Which methods are buyers actually comparing?

The shortlist is stable. Promo codes, vanity URLS, post-purchase surveys, publisher-reported delivery data, matched-market tests and holdout or incrementality tests cover most campaigns.

Each sits somewhere on two axes: cost per measurement and strength of causal claim. Cheap methods tend to measure correlation. Expensive ones try to isolate cause.

The measurement and reporting guide sets out how these methods fit into a wider reporting cycle, which is worth reading before you commit budget to any single approach.

How do promo codes and vanity URLS work?

A host reads a code, listeners enter it at checkout, and you count redemptions. It is simple, cheap and easy to explain to a finance team.

It also undercounts. Most listeners do not use a code even when they buy, and codes cannot capture brand effects or later purchases.

In the UK, typical code redemption rates for podcast ads are often in the low single digits, usually around 1% to 5% of listeners. The exact figure depends on the offer and checkout friction. If a code records 40 redemptions, treat that as a labelled illustrative example of minimum direct response, not the campaign's full contribution.

What do platform delivery figures tell you?

Download counts, reach estimates and completion rates come from hosting platforms and podcast networks. They describe how many people received the ad and how much of it they heard.

They say nothing about sales on their own. A campaign can deliver millions of impressions and shift no revenue.

That does not make them useless. Delivery data checks whether a campaign ran as booked, which supports any later performance claim you publish.

Where do surveys and matched-market tests fit?

Post-purchase surveys ask buyers how they heard about you. They capture responses codes miss, including brand-led purchases.

Matched-market tests compare regions or periods with and without the campaign. They are stronger on causation but need enough volume to detect an effect.

Both approaches rely on personal data at some stage, so the ICO's direct marketing guidance on targeting and list use is relevant to how you collect and hold responses.

What does incrementality testing add?

Incrementality testing splits your audience into exposed and unexposed groups and compares outcomes. The difference is your estimated lift.

It is the closest thing to a causal answer, and the most demanding. You need sufficient scale, clean assignment and patience.

Run it on your largest campaigns, not every episode. Smaller buys rarely produce a readable result.

How do the methods compare?

Relative cost means the typical budget and effort needed to run the method, from low for codes to high for holdout tests. Speed means how quickly you get a usable result, from fast to slow.

What compliance points apply to attribution claims?

Any performance figure you publish is a commercial claim. The UK advertising self-regulation system sets out who enforces the CAP Code and how complaints are handled.

Misleading claims about results can also fall foul of consumer protection law. The Consumer Protection from Unfair Trading Regulations 2008 prohibit unfair commercial practices, including misleading podcast ads, while the Climate Change Act 2008 is not that law. The Digital Markets, Competition and Consumers Act 2024 updates the framework.

Keep the method behind every number on file. If a code undercounts, say so in the report rather than presenting it as total return.

A reporting dashboard helps here, because it keeps method notes alongside the figures instead of separating them.

Common questions

Can I rely on promo codes alone?

No. Codes capture only listeners who choose to use them, so they understate total response. Use them alongside at least one other method.

Which method is best for a small budget?

Start with codes or vanity URLS plus platform delivery data. Add a post-purchase survey once volume justifies the extra work.

How often should I run an incrementality test?

Once or twice a year on your largest campaigns is usually enough. More frequent testing rarely repays the cost on smaller buys.

Do delivery figures count as attribution?

Not on their own. They show the ad ran and was heard, which is delivery evidence rather than proof of sales effect.