Wella's Nioxin Pro Clinical line generated action from 70 percent of exposed listeners in an audio advertising campaign built on artificial intelligence, according to results published today by Azerion and media agency Digitas.

The campaign, delivered for Wella's hair-thinning and scalp-health product line, used AI-generated personalised dynamic creative rather than a single fixed script. Azerion built the audio through its agentic production platform, Audiostack AI, generating 100 distinct versions tailored to specific audience segments and locations. The result, according to a brand lift study, was a campaign that moved more listeners to act and left a stronger impression than a generic approach would have.

What the brand lift study found

The study measured several metrics comparing people exposed to the campaign against a control group who had not heard the ads. According to the announcement, 70 percent of exposed listeners said they had taken, or would take, some form of action after hearing an advertisement. That action could mean buying the product, researching it further, or passing details to someone else. Among people who reported having suffered from hair thinning themselves, the effects went further: ad recall rose by 21 points and awareness increased by 22 points, relative to the control group.

Clarity of branding also scored well. Seventy-seven percent of exposed listeners said the branding in the ad was clear to them, a figure that speaks to a persistent worry in audio advertising: that listeners, unlike viewers of a banner or video ad, cannot glance back at a static visual to confirm which brand is speaking to them. A radio or streaming spot has one chance, in real time, to make the sponsor unmistakable.

The measurement itself came from On Device Research, which structured the study as an exposed-versus-control comparison. According to the source materials, the design tested location-specific store ads against direct-to-online versions of the same creative. Listeners within a defined radius of a participating Boots store heard a version directing them to that outlet; those outside the radius heard a version pointing to Boots.com instead. People who had been exposed to the campaign outside a controlled testing environment were later retargeted with a brand lift survey, distinguishing the exposed group from the unexposed control.

How the creative was built

Azerion's approach rested on segmentation along two axes: audience and geography. The audience segments were defined as women between 25 and 45 and men between 20 and 35. Layered onto that demographic split were 24 specific UK locations, each of which received its own location-referencing version of the ad. Listeners outside those 24 areas were not left with nothing; instead, they heard a more generic version of the creative that pointed toward the websites of retailers stocking Nioxin products, specifically Boots and Lookfantastic.

In practice, this meant a listener's experience of the ad depended on where they happened to be when it played. Someone within one of the 24 target regions heard a version referencing their locality, nudging them toward a nearby retail point. Someone elsewhere heard a version built around the same underlying message but pointed at an online retailer instead. Building that many permutations by hand, across two demographic groups and two dozen locations plus a fallback version, would traditionally have required substantially more production time and cost than a single studio recording. Audiostack AI, according to the companies, allowed Azerion to generate the full set of 100 versions without that overhead.

Why hair thinning required a different approach

Both agencies involved in the campaign described hair thinning as a category that resists generic messaging. Ruth Reynolds, insight and strategy director for the UK at Azerion, said the personalisation strategy mattered specifically because "hair thinning can be a sensitive topic, and different for each person that experiences it." That sensitivity, according to Reynolds, is why relevance at the moment of listening carried weight beyond a simple reach or frequency metric: "Ads were relevant to the person hearing them at that particular moment which drove a significant number of listeners to act on the message, while raising brand awareness among a varied target audience."

Julia Drucker-Allister, EMEA media director at Wella, framed the results in terms of a shift away from undifferentiated advertising. "This campaign represents a significant step forward in how we use technology to better serve our consumers, away from a 'one size fits all approach' and into personalisation at scale," she said. Drucker-Allister connected the recall figures directly to the localisation strategy, adding that "when we make our messaging more localised and relevant, we don't just build awareness - we build a genuine connection with our audience."

Lisa McAllister, group media director at Digitas, described the underlying production challenge as one of matching creative output to a data-driven media strategy without losing emotional nuance. "By partnering with Azerion to build an AI-powered audio solution, we were able to deploy 100 hyper-relevant variants that responded to the listener's context in real-time," McAllister said. She characterised the resulting metrics as evidence of a broader shift in how personalisation should be regarded within the industry: "The double-digit lift in brand awareness proves that personalisation is no longer just a luxury - it's the new benchmark for effective audio."

The research behind the campaign

The Nioxin results did not emerge in isolation. They follow directly from a separate study Azerion published alongside Differentology, a market research consultancy, in a white paper titled "The Voice of the Future: What the Latest Research Tells Us About AI Audio Advertising." PPC Land reported on that research when it was released on July 14, 2026, detailing how 3,000 respondents across the United Kingdom were exposed to different combinations of human and AI-generated voiceover across two established retail brands.

That earlier study tested neutral human voice against three separate AI-driven conditions, including a regional AI voice condition using four distinct accents and a Dynamic Creative Optimisation condition built around personalisation. According to the earlier white paper, the central finding was that AI-generated voice ads performed comparably to, and in some measures better than, standard human voiceovers, despite a substantial share of listeners assuming beforehand that a human voice would prove more persuasive. The Nioxin campaign appears to operationalise that earlier finding in a live commercial deployment, applying the DCO methodology to a specific product line rather than a controlled research environment.

Dynamic Creative Optimisation itself is not a new concept in advertising broadly. The technique, which assembles ad components in real time according to the specific needs of an impression, has existed across display and video formats for years. What has changed, particularly in audio, is the ease with which a large volume of distinct variants can be produced. Where DCO in display advertising typically swaps images, headlines, or calls to action, audio DCO must generate distinct voice recordings, a task that historically required either a large roster of voice actors or accepting a smaller number of generic variants. AI voice generation removes much of that constraint, at least for advertisers willing to use synthetic voice technology in place of, or alongside, human talent.

Audio's persistent measurement gap

The Nioxin campaign lands within a wider structural pattern that PPC Land has tracked across the audio advertising sector. Earlier PPC Land coverage documented that consumers dedicate roughly 31 percent of their media time to audio platforms, while advertisers allocate only about 9 percent of ad budgets to the channel, a 22-percentage-point gap that has persisted despite improvements in programmatic infrastructure and measurement standardisation. That imbalance has been attributed, in various industry discussions PPC Land has followed, to both a plumbing problem, in which different platforms report different things and make cross-platform comparison difficult, and a positioning problem, in which audio has not been marketed to buyers with the same urgency as video or display formats.

Brand lift measurement specifically has been moving toward faster, more integrated delivery across the wider advertising industry. PPC Land's coverage of a June 2026 partnership between Basis and Cint noted that a January 2026 partnership between DAX and DISQO had already brought brand lift measurement directly into audio campaigns, extending a pattern of embedding measurement tools inside the platforms where media is bought rather than treating brand lift as a separate, standalone research exercise. The consistent thread across these announcements, according to that PPC Land analysis, is that measurement which once operated apart from campaign execution is increasingly built into the platforms handling activation itself.

Separately, the State of Audio Adtech Report 2025 from AdsWizz found a comparable 22 percent gap between audio engagement and advertising investment, a figure PPC Land has referenced across multiple articles addressing the format's persistent undervaluation relative to consumption time.

AI creative adoption across the wider industry

The Nioxin campaign also sits inside a broader industry movement toward AI-assisted creative production, though evidence on quality outcomes remains mixed. A WARC and TikTok study published on July 14, 2026 found that 88 percent of marketers reported higher creative volume since adopting generative AI tools, yet only 45 percent reported a significant improvement in creative quality. That research, drawn from 400 marketers across the UK, US, Australia, and Brazil, suggests that volume gains from AI-assisted production do not automatically translate into stronger creative performance, a distinction that makes brand lift results, rather than output volume alone, a more meaningful signal of whether a given AI-driven campaign succeeded.

Separately, an IAB study covered by PPC Land in July 2025 projected that generative AI creative would account for 40 percent of all advertisements by 2026, with 86 percent of buyers surveyed already using or planning to use generative AI for video ad creative specifically. The Nioxin campaign extends that broader trend into the audio channel, applying similar logic, personalised, AI-generated variants at a volume manual production could not match, to a category the companies involved describe as requiring particular sensitivity.

Company context

Azerion, listed on Euronext Amsterdam under the ticker AZRN, was founded in 2014 and is headquartered in Amsterdam, with commercial teams operating in more than 17 countries. The company has built out a series of audio and omnichannel supply partnerships over the past year. PPC Land reported in May 2026 that Azerion integrated the Spotify Ad Exchange directly into its Hawk demand-side platform on May 28, 2026, giving advertisers programmatic access to Spotify's audio, video, display, and podcast inventory without an intermediary connection. Earlier that same month, Azerion took over programmatic monetisation of digital audio advertising across roughly fifteen L'Equipe sports podcasts in France through an agreement with Amaury Media, dated May 4, 2026.

Digitas, the media agency that partnered with Azerion on the Nioxin campaign, operates as a subsidiary of Publicis Groupe, the France-based communications holding company. Publicis has been expanding its data and measurement capabilities through acquisition over the same period; PPC Land reported in May 2026 that Publicis agreed to acquire LiveRamp for 2.5 billion dollars, a deal framed around agentic AI data capabilities across the agency network.

Numbers behind the campaign

Several figures anchor the results Wella and its partners published. The headline action metric, 70 percent, describes listeners who reported taking or intending to take some form of action after exposure, a broad category spanning purchase intent, further research, and word-of-mouth referral. The 77 percent branding-clarity figure addresses a distinct question: not whether listeners acted, but whether they correctly identified the advertiser at all, a metric of particular relevance in audio, where no visual reinforcement exists to correct a misattributed message.

The 21-point recall increase and 22-point awareness increase both apply specifically to the subset of listeners who reported having personally experienced hair thinning, rather than to the exposed audience as a whole. That distinction matters for interpreting the results: the campaign's strongest documented effects concentrate among people with a pre-existing, personal connection to the product category, a pattern consistent with the sensitivity both Reynolds and Drucker-Allister described in their comments on the campaign's design.

The creative volume figure, 100 versions, reflects the combination of two audience segments and 24 specific locations, plus generic retailer-directed fallback versions for listeners outside the 24 target areas. No overall reach, spend, or campaign duration figures accompanied the published results.

Timeline

Summary

Who: Wella, the personal care brand behind Nioxin Pro Clinical, worked with omnichannel advertising platform Azerion and media agency Digitas to deliver the campaign. On Device Research conducted the brand lift measurement.

What: An AI-generated, personalised audio advertising campaign for Nioxin Pro Clinical, a hair-thinning and scalp-health product line, produced 100 distinct creative variants using Azerion's Audiostack AI platform. A brand lift study found that 70 percent of exposed listeners took or intended to take action after hearing an advertisement, with recall rising 21 points and awareness rising 22 points among people who had personally experienced hair thinning.

When: The results were published today, July 29, 2026. The underlying research methodology draws on a separate Azerion and Differentology study released on July 14, 2026.

Where: The campaign targeted listeners across the United Kingdom, using 24 specific locations for geographically tailored creative and directing other listeners to online retailers Boots and Lookfantastic.

Why: The results matter to the marketing community because they apply a documented industry finding, that AI-generated voice can match or exceed human voiceover effectiveness, to a live commercial campaign in a sensitive product category. The campaign also arrives as brand lift measurement moves increasingly into real-time, platform-embedded tools, and as the wider industry grapples with a persistent gap between rising AI creative volume and inconsistent gains in creative quality.