Optimising digital investment and media efficiency for luxury in China

It is worth looking at how the relationship with luxury is changing in China as this will, in some part, determine the digital investments and media channels that will be most important for the immediate future.

Luxury has always had a strong 'experiential' element to it – brands have drawn audiences into relationships through exceptional service, exclusive events and more. What is particularly interesting though, is how this audience can work at a global rather than a national level; they are not constrained by geography and are often interested to interact with their peers across the world.

Kantar research carried out in China, Europe and N.America shows that there are shared interests in discovery, exclusivity and status that are particularly pertinent to media strategy. What is different today is the nuances and links with other parts of culture (music, literature, cult TV) and how this throws down some challenges for luxury brands.

What does this mean for media strategy?

Firstly, digital channels have had an image of being the low-cost / more-youthful / mass-market route to audiences – but what if we flip this perception and argue that the luxury audience values digital connection because it recognises their desire to be productive and efficient? However, Bain research states that most luxury categories in China still have low online penetration rates.

Secondly, luxury brands should acknowledge and even honour this audience's sense of status – because digital channels can offer a rich experience that aids discovery, personal attention, control, and connection to peers with a similar status. HNWIs trust other HNWIs, prefer discrete customer service and are best reached through private social domains and closed social channels, as they are very protective of who they are and what they do.

These two points mirror a bigger point about digital now working across brand / performance investments.

Investment proportion of brand VS performance media? 

In China, in comparison with other parts of Asia, the purchase research phase of luxury brands skews heavily towards digital (especially on Tencent-owned WeChat) but the actual purchase remains offline / in-store and via owned media for the majority. This implies the balance of investment should be towards brand media, ensuring that mental availability and brand standout remains high within what can be very cluttered environments in China.

There are a couple of issues that run counter to this principle though. The interest in 'self-gifting' may translate powerfully into 'performance' strategies – the final nudge to indulge for oneself. Also, China's plan to turn Hainan into the world's largest free-trade port by 2025 is predicted to change the pricing landscape for international luxury brands. After the customs closure, brands will be able to sell directly to their end consumers. This will give brands a much better understanding of buying habits, previously limited by the grey market in China.

In general, luxury brands are rarely advised to shift spend into performance strategies, however, with the rise of 'end to end' digital commerce there is a greater opportunity to rebalance spend a little.

Even during high pressure sales events, investment in brand is still the powerful route to take. The importance of the big sales days inevitably puts attention onto performance spend but this may not actually be the best approach for luxury brands, shown through a case study from cosmetics giant L'Oreal in China.

In arguably the most competitive and valuable e-commerce event in the world, it is a brave brand that steps away from the standard focus on performance strategy and opts for brand building. What we want to highlight in this work is the way that L'Oreal Paris recognised the importance of running a full campaign strategy rather than a 'digital execution' around the event in China.

The campaign had a full thirty-day strategy to draw the luxury audience into the detail of La Maison, its storytelling and opportunity to engage. The final part of the puzzle was then to link into the Singles Day event and draw the audience from the clutter of the big platforms, to become a host rather than a 'fighter' for attention. Rather than going to the big e-commerce platforms as a guest, the brand wanted to create its own Double 11 platform, to siphon Double 11's consumers to its own platform as L'Oréal Paris consumers.

How should luxury brands optimize media efficiency? 

Efficiency is a high-risk mantra for luxury brands.

The luxury audience demands experience, innovation, leadership – the ability to build partnerships, to quickly adopt new opportunities is paramount. Tmall Luxury Pavilion and Farfetch in China, for example, have differing traffic volumes, commission rates, GMV contributions, MAUs, consumer profiling, average transaction values, and purchase frequencies.

We are going to use this section to segue into measurement and to show that media efficiency needs to match the specific need of the brand. Pernod Ricard shows in this case study that a multi-disciplinary team can guide media strategy by product line very efficiently (and effectively).

For example, its Lillet brand, the specialty aperitif, was vastly ambitious to build awareness in the upper funnel (A1 in Bytedance’s 4A model). 8 video assets were mapped against 19 audience segments, on aggregate there were 152 sets of 'creative + audience' combinations constituted in the first place. Activated via Douyin's biddable feed ads, all these combinations were tested in two weeks. And 3 out of 152 combinations with the best performance were allocated more media investment.

The media objective for the Chivas Regal brand was slightly different, putting more emphasis on mid-lower funnel conversion i.e. acquisition of A3 (Consideration) and expansion of A4 (Purchase) audiences. Building on top of ByteDance’s advanced CDP Yuntu, the team fully utilised a variety of media tactics and placements including more precise targeting with higher frequency.

Is MMM (marketing mix modeling) workable for luxury brands?

MMM works if the luxury brand has enough high-quality data to guide it as MMM needs at least two years of high-quality data to be input. Luxury brands should ensure that they are focused on what is most critical to them. They are not searching for reach / awareness. They may be better advised to look at the specifics of brand health, innovation and relevance indexes for their TA, rather than the advertising performance as a whole.

In our WARC report "The Future of Measurement", we put forward the view that the MMM models struggle to reflect how retail media has risen in importance – but that the innovation in modelling offers brands cost-effective ways to understand and guide their strategy.

Luxury brands need a disproportionate focus on leadership, difference and experience. This is possible to pick up through MMM but the model would not be the 'off the shelf' options so frequently discussed today.

What is the optimal allocation in terms of online and offline adspend? 

Analytic Partners‘ ROI Genome research makes a strong case for a combined multi-channel approach to drive campaign effectiveness. For example, combining offline and online media channels is 45% more efficient than offline alone or online alone.

Here, we see it as a broader issue of deploying an adspend strategy that is experiential, innovative and appropriate for brands seeking to be seen as leaders. Going offline as others rush in to digital provides this, but we also see how Burberry prospered through partnerships – so, it is the innovation rather than channel type that is key.

When Burberry first partnered with Instagram & WeChat for instant online buys of its Thomas Burberry Monogram collection in 2018, the brand was significantly rejuvenated, arguably quite fundamentally repositioned – so it went and did it again with Twitch for Burberry Spring/Summer 2021 show livestreaming.

A recent report by Tencent estimates that 90% of luxury purchase decisions, regardless of channel, are influenced by online information. And themes of personal benefit and exclusivity are important for brand advocacy amongst HNWIs. When it comes to promoting their favourite brand online, they are more likely to do so (source: CNN/Ipsos) because doing so will enhance their own reputation or status. It shows they have more in-depth knowledge or have been given exclusive access to content or services.

Thus, luxury brands are urged to focus on opportunities to deliver standout experiences and innovation leadership – of course, which may also be offline. For some sub-categories in luxury (like jewellery), the offline opportunity is particularly important. Despite many luxury brands being 'ahead of the game' with digital commerce, they understood that a 'physical' experience carried the highest value.

Therefore, the decision to move adspend across channels is one guided by what can deliver the greatest experience' rather than the mass-market brand need for ‘reach’ and ‘attention’. One caveat: luxury brands must take care to maintain a high-end experience that is on-brand when buying online media. Louis Vuitton’s livestream debut on Little Red Book in 2020 drew mixed responses as some felt the execution of the livestream was not premium enough.

Lastly, experience can also be thought of as 'exclusive inclusion' and the media strategy should burnish that balance of allocation, online or offline.

Given stricter regulations in China, what to double down on?

For the luxury market, dataset building and digital transformation is still in its early stages, even in the Chinese context.

In China right now, more than 90% of traffic is mobile traffic. It's quite clear about what we could collect in terms of device IDs (Open Anonymous Device ID A device ID used in China instead of GAID or IMEI/MEID) or UnionIDs on WeChat, but with stricter government censorship and tighter data privacy, it is hard to collect fuller data.

Globally, you have Google, Amazon, Meta, Microsoft, Apple (GAMMA) but in China, you have JD, Alibaba, Tencent, Bytedance – all big tech ecosystem platforms that don't 'talk' to each other. All the data actually stays with their ecosystems. If you want to lead audiences from Tencent to Alibaba, it is difficult.

A key focus for luxury brands should be on collecting better first-party data to understand more about consumer journeys, and eventually to achieve customized or personalized communication with luxury audiences for each single touchpoint.

An important project Coty is doing in China is to set up its own data infrastructure: collecting own first-party data, leveraging data pools, working with other data providers (like China Unionpay that has a lot of consumption behaviour data), mapping audience profiles, running digital campaigns, and then comparing performance with control groups that doesn't have their own tagging.

It is a way to collaborate with the big ecosystems and further build on their data. "In the short term, you have no choice. You have to work with these big ecosystems that would not let you have 100% control of the data. Otherwise, they lose their advantages if you got the data," said Alan Zhao, Coty Inc's media and data director in a WARC interview. 

Zhao mentioned that ad fraud is a concern in China, especially for brands working with third-party DSPs. However, Coty China mitigates this risk by working directly with big ecosystems like Tencent and Alibaba. He sees a trend of more ad dollars going into big ecosystems and away from open exchanges. 

What luxury brands can try to do is to leverage big tech’s data capabilities to set up their own pools to be as big as possible, identify the most effective audience profiles and find similar lookalikes in the big ecosystems – and then acquisition costs of new audiences will be lower in theory. 

Apart from cost efficiency, what adtech players need to prove to their clients is the outcome of driving online audiences to offline stores, and vice versa. If clients spend two million, is it bringing 10k audience to their offline flagship stores? If they collect the data of people who has seen offline ads on premium screens, and then uploaded to the big ecosystems to be matched, do audiences being bought online actually have higher consumption power?

"So, that would be the future. Because in China, most luxury brands have already touched/reached all the audiences that have the potential to buy a product. The next step is to really cultivate existing audiences to become a loyal member of the brand to increase the repeat purchase rate," Zhao pointed out. 

First-party data is core to media strategy to win in China’s current digital age. But it depends on how a luxury brand sets up its entire data infrastructure. There are a lot of opportunities if you fully merge with all possible private domain touchpoints like Enterprise WeChat, one-to-one BA (beauty advisor) communications (usually unrecorded and unstructured), and social accounts of your own brands, etc. From the digital perspective, it makes sense; on the business perspective, it must be top down with the CEO pushing it.

Last but not least, reputation is so fundamental to luxury brands that from an effectiveness perspective, a misstep is likely to hold more risk than there is potential for reward. Avoid the risks brought by automated decision-making that cause data-enabled price discrimination & misleading AI recommendations.