The Performance Measurement Gap: Why Marketers Need to Connect Media Spend to Revenue

By Dana Crow, US Managing Director, Client Success at IDX

Performance marketing has a measurement problem, and it’s costing brands money as US advertisers are paying more for media while getting less visibility into what their investment delivers. Privacy changes have weakened traditional tracking, customer journeys now span more channels, and AI answer engines are changing how people discover brands before they ever reach a website.

Yet too many marketers are still relying on last-click attribution, platform-reported ROAS and lead volumes to prove performance. These metrics might look impressive on a dashboard, but how much do they really tell us about what’s driving business growth?

When advertising platforms are effectively grading their own homework, it’s difficult to know which campaigns are genuinely generating new business and which are simply taking credit for conversions that would have happened anyway.

The uncomfortable truth is that much of the industry is still measuring activity rather than results.

Cheap leads don’t necessarily mean good business

For years, performance marketing has been built around driving clicks, generating leads and bringing down acquisition costs. The logic was straightforward: make campaigns more efficient and the business will benefit.

But that only works if those leads become paying customers. Take a senior living provider, for example: someone might click an ad, complete an enquiry form and then spend six months researching options, visiting communities and discussing the decision with family. By the time they move in, the original advertising interaction is a distant memory.

A dashboard celebrating a low cost per lead tells us very little about whether that campaign was commercially successful, and the same applies to higher education and financial services industries too, where the difference between an initial enquiry and a completed enrollment or funded account can be substantial.

For these businesses, cost per move-in, cost per enrollment and cost per funded account are far more valuable measures of success. While lead volume still very much matters, a campaign generating 100 cheap enquiries that never convert isn’t necessarily performing better than one delivering 20 qualified prospects who become customers. Marketers need to know the difference before deciding where to spend their next dollar.

Stop letting platforms grade their own homework

There’s another problem with traditional measurement: too much of it relies on advertising platforms reporting their own success.

Google, Meta and other major platforms provide sophisticated reporting tools, but each sees only part of the customer journey. Multiple platforms can claim credit for the same conversion, while activity happening offline may never appear in their reports – making it hardly surprising that brands are questioning whether those numbers reflect genuine growth.

This is where first-party data becomes so important. Connecting advertising platforms with customer relationship management (CRM) systems allows marketers to look beyond the initial conversion and understand what happens next. Which leads become customers? How long does that take? And which campaigns are attracting people who actually generate revenue? That information can change investment decisions considerably.

A channel that looks expensive based on lead costs might deliver the strongest commercial returns, while another that appears highly efficient could be filling the sales pipeline with prospects who never buy. However, without connecting media data to business outcomes, marketers are making budget decisions with only half the picture.

There’s no perfect attribution model

Of course, better first-party data doesn’t mean every sale can be traced neatly back to an ad. The industry has spent years searching for a single source of truth, but customer behaviour doesn’t work that way. Instead, marketers need to get comfortable using different measurement approaches together.

Media mix modelling can help identify how different channels contribute to overall business performance. Incremental testing can show whether advertising is generating additional sales or simply claiming credit for conversions that would have happened anyway, and attribution can still provide useful insights into individual interactions. Neither is perfect or particularly useful on its own; however, when used together, they offer a much stronger basis for deciding where budgets should go. And when the results disagree, that’s worth investigating rather than ignoring.

AI won’t fix bad measurement

AI is making campaign management faster, from automated bidding and targeting to generating and testing creative variations. It can also help marketers clean data, identify performance changes and make sense of complex reporting without spending hours digging through spreadsheets, which can be really useful. However, there’s a danger in assuming more automation automatically means better performance. If an AI-powered campaign is optimizing toward cheap leads rather than customers who generate revenue, it could simply help a brand waste money faster.

The same goes for creative. Generative AI might allow a team to test 20 variations instead of two, but producing more ads isn’t the same as producing effective ones. Marketers still need to understand which messages influence customer decisions and why. AI can guide strategy, but it shouldn’t define it – the quality of the inputs and the judgment behind the decisions still hold the key to a successful campaign.

The next dollar matters more than the last click

As media costs rise, the pressure to demonstrate marketing’s commercial value will only intensify, and that means changing the conversation from how many clicks a campaign generated to how much business it actually brought in.

Performance teams need to connect their media investment to CRM data, test whether campaigns are delivering incremental growth and use those findings to make better decisions. After all, a strong ROAS figure might impress in a marketing meeting, but if it doesn’t reflect the revenue advertising actually contributed, it’s a number that doesn’t mean very much at all.