From validating decisions to shaping strategy

Historically, measurement has largely been treated as marketing’s final step. A campaign launches, the numbers come in and someone is asked to explain if it drove growth and if not, why. It has been considered as a reporting and dashboard-building function for a while, especially since using platform-attributed performance data. A way of validating decisions that have already been made rather than a strategic approach that can influence the media plan. That model is now quickly becoming outdated, especially as relying on attributed data will stop once cookies are phased out. The most interesting shift happening in performance marketing today isn’t just AI, attribution or even the decline of third-party cookies. It’s that measurement is moving upstream in the decision-making process. Increasingly, it’s shaping strategy before a single pound of media budget is spent.

What AI actually means for the role of measurement

It’s tempting to frame every conversation around AI because it’s impossible to ignore. New search behaviours, large language models and conversational interfaces are changing how people discover brands.

However, the more meaningful question is what those changes mean for the role of measurement. Obviously, AI will help automate the analysis and provide a quicker, more detailed view of the campaign performance, but the answer, in my experience, is that measurement is becoming less about reporting campaign performance and more about helping businesses improve their buying strategy and have a more informed media plan and ensuring everything we do is measurable and actually makes sense for the business.

Case Study: Hyper-local Out of Home (OOH)

A recent example would be a hyper-local OOH for an app-first client targeting a niche audience. Measuring an activation like this demands a radical departure from broad-brush analytics. We weren’t buying mega-screens at Piccadilly Circus; we were placing micro-formats in corner shops and pubs. Converts don’t download an app over a pint. They do it at home, two tube stops away. To capture this displaced journey, our MMP had to measure incrementality at the postcode sector level, mapping both target venues and their surrounding commuter belts. Crucially, if an area couldn’t be measured with this degree of precision, it was ruthlessly pruned from the media plan. By anchoring these localised cohorts to first-party data, we can accurately track true LTV over months or years. If you can’t prove a campaign’s impact with that level of granularity, ask your CFO a simple question before spending a penny: is it even worth running?

Why brands have invested heavily in 1PD and CRM

This shift has been building for years. As privacy regulations have tightened and third-party data have become less available, brands have invested heavily in 1PD and CRM systems. Five years ago, CRM was not seen as a priority, but today both agencies and brands are putting it at the forefront of their strategy. Most have rich customer datasets containing purchase history, browsing behaviour, preferences and engagement across multiple channels. Collecting that information is less of a challenge. The real challenge is knowing what to do with it, in a world where ad personalisation is no longer a pleasant surprise; it’s the expectation. Broad audience segments and average campaign performance are still useful, but they don’t tell the full story. Brands increasingly need to understand why their campaign worked, who it worked for and how they can repeat that success.

Marketing insights that become commercial insights

That’s where measurement starts to become genuinely strategic. Traditionally, we measure performance at campaign or channel level. Did paid search outperform social? Which creative generated the strongest return? Those questions remain important, but today’s technology allows us to go much deeper. For example, instead of simply knowing that a footwear campaign drove incremental sales, we can begin to understand which products resonated with which audiences, and why. If one customer consistently responds to new collection trainers while another converts only when shown a reminder about their favourite design, those aren’t just marketing insights. They’re commercial insights.

Measurement is informing all departments

That changes the strategic conversation and the role of media measurement inside media agencies as well as brands. Measurement is informing all departments – from sales and R&D through merchandising decisions and product development, to unquestionably marketing via creative strategy and media planning. If brands can measure performance at a much more granular level, they can build products, experiences and campaigns around evidence and consumer feedback rather than instinct.

Measurement at ROAST

At ROAST, our client conversations reflect a clear shift: measurement is now treated with the same strategic priority as paid media and SEO, serving as a genuine competitive advantage. The focus has moved beyond simple transactional returns – spending £100 on Meta expecting £150 back – to understanding precisely where media spend is maximised. Robust measurement provides the clarity needed to optimise cross-channel budgets and align media plans directly with verifiable outcomes.

Incremental value across an increasingly fragmented journey

Even the emergence of AI-powered search reinforces this trend. Whether a customer starts with a search engine, including GEO, a social platform or an OOH billboard is less important than understanding the incremental value each channel creates, and how they work in conjunction. The customer journey has never been linear, and it is becoming even less visible. The challenge isn’t capturing every touchpoint perfectly, which we know is impossible. It’s understanding which investments genuinely create growth and which simply move customers around an increasingly fragmented journey.

From telling you what happened to determining what happens tomorrow

That’s why the future of measurement isn’t about building more dashboards or more complex attribution models (which will disappear alongside cookies anyway). It’s about helping businesses make better decisions. The organisations that win over the next few years won’t necessarily be the ones with the most sophisticated technology or the biggest datasets. They’ll be the ones that use measurement to influence strategy rather than simply evaluate it.

For a long time, measurement sat quietly in the background of marketing. That era is ending. It’s no longer just the function that tells you what happened yesterday. It’s becoming the discipline that helps determine what happens tomorrow and that’s a much more interesting role for performance marketing to play.

A version of this article was featured in Performance Marketing World. 

From August 2026, Google will use IP addresses for ad measurement and personalisation in the UK, EEA and Switzerland, triggering new consent requirements.

Here’s what it means for brands.

What the update is:

• Google is now expanding how it’s allowed to use IP addresses across the UK, EEA and Switzerland, moving away from purely operational uses (serving ads, routing traffic, fraud prevention etc.) to ads measurement and ads personalisation – Something the rest of the world mostly has been doing already and we’re catching up.
• Google isn’t collecting new data – it already receives IP addresses however the key change is declaring the purpose of this data.
• Naturally IP addresses count as personal data under UK/EU GDPR, this now triggers consent requirements that weren’t previously in required. Hence sites will be required to run a certified Consent Management Platform (CMP) to serve ads to EEA/UK/Swiss users going forward – with ongoing audits to check compliance.

What Does This Mean for Brands Operating across the UK, EEA and Switzerland

New measurement signal

IP-based identification gives another way to stitch together cross-device attribution and measure real-world outcomes (e.g., store visits) as cookies and other identifiers degrade. Whilst the use of UUID’s is the more optimal setup, this serves as a good interim solution.

Consent becomes the gating factor

Businesses that don’t have a CMP correctly setup risk being unable to access these new measurement capabilities – meaning a compliance gap could directly translate into lost ad performance or revenue. This is a strong prompt to audit your CMP setups now (are they TCF-certified, is Feature 3 consent being properly requested and passed?)

Privacy scrutiny likely to increase

Unlike cookies, IP addresses can’t be deleted by users, and the ICO have flagged concerns about persistent device recognition, so we can expect continued regulatory attention and potential brand nervousness around this, worth getting ahead of in any consent-related conversations.