Why agencies in Asia Pacific are starting to embrace the science of data
It is already a common sight to see data scientists working in advertising and marketing technology companies like Facebook, and in traditional media companies like BBC, Financial Times and Sky Media.
As data-driven marketing becomes increasingly important for brands to reach their right target audience in recent times, creative, digital media, mobile and marketing agencies in Asia Pacific are starting to hire more data scientists, build in-house data science teams and acquire companies that specialise in data science.
In addition, data science companies are setting up shop in APAC to meet the growing demand to maximise in-house data production and analysis. In November, French software developer Dataiku Inc., which owns predictive analytics software platform Data Science Studio, announced its intention to accelerate its expansion in this region.
Brands like Cisco Systems also see the value in having in-house data science teams because it is moving more of its marketing to digital, Mark Phibbs, vice president of marketing and communications for APAC and Japan at Cisco tells The Drum.
“We have already invested in analytics and data science to analyse share of wallet and market opportunity,” says the Australian. “We are now investing in what we call market insights and operations, in areas including: technology adoption; database operations; testing and optimisation; analytics; market insights; and strategic planning.”
For agencies, data science is quickly becoming a key aspect of their services that they offer to clients because it enables the necessary marketing transformation component of allowing consumers to engage with brands in a completely omni-channel way, as part of a journey towards completely personalised marketing, says Matt Sutton, chief executive, Asia Pacific at AdParlor.
“From an execution perspective this means all brands need to be using data science to create a single unified view of their own consumer’s journey and at the same time amalgamating that with third-party data on what those consumers and other potential consumers are doing, where this places them on the marketing funnel, and as such what messages to be delivering to them.”
“In this world, any agency partner for a brand needs to have a deep understanding of data modelling and be able to craft specific messages for specific consumer behaviour. You can’t play an agency role and not offer data science services and capabilities,” he adds.
As clients want strong insights and more evidence to support great creativity, in addition to more data being collected and organise, it would be ‘silly’ to just ignore data science, opines Oliver Eriksson, regional global advisory head at VML Southeast Asia and India.
Eriksson however, points out that the real challenge for agencies is getting access to that data. “If you don’t have a strong partnership with your client, it’s going to be difficult to get hold of meaningful amounts of data or at least have a strong enough partnership to be able to collaborate with their internal data team.”
“At the same time, it’s important to keep in mind that awesome data analysis doesn’t necessarily tell you what to do about it. Just as critical would be upskilling the planning team to create a stronger partnership between data scientists and planners, so together they can obtain a better understanding of the audience or results of campaigns, and leverage that understanding to drive smarter marketing for brands,” he adds.
Daniel Hughes, senior vice president and chief data officer at DigitasLBi, agrees with Eriksson, and adds that data science is critical for agencies because as clients are under immense pressure to be accountable for every investment and decision, accountability flows downhill to their agency partners.
“There are no excuses to hide behind. Data is abundant because consumer interactions leave digital trails. The analysis tools are inexpensive thanks to cloud computing and open source,” he explains.
Explaining why agencies are starting to maximise in-house data production and analysis in APAC, Nicole Liebmann, head of mobile for APAC and South Africa at Exponential says that data science allows agencies to properly understand audiences and the media channels they dwell in.
“This then enables them to identify the best ways to utilise specific channels to ensure effective communication that drives engagement and action,” she says. “We need to shift from just driving reach; it isn’t about ‘spraying and praying’ and hoping that the mass exposure will translate into success for brands. Rather, it is about intelligently buying impressions or using data science to build out campaign elements.”
Liebmann also highlights that doing so allows the campaign to reach a relevant audience in a meaningful and effective way, ultimately driving the best possible results. Embracing data science also allows agencies to become a more strategic partner for brands who may have their own data, but are not able to make that data actionable in media, she adds.
For AdParlor’s Sutton, having an in-house data science team is not a nice to have, it is a ‘basic hygiene factor’. “All advertisers need to be adapting for a world where all marketing is programmatic, and first- and third-party data comes together and enables the crafting of messages based on consumer behaviour and purchase intent,”
“As such, building these competencies will become an existential challenge for brands. The agency’s role is always to enable an advertiser for marketing success and this requires both technology and human resource commitment to data science,” he adds.
Even though there are many benefits for agencies to have in-house data science teams, Boomi Boominathan, an Indian-born data scientist who manages a Philippines-based team for Italian digital advertising company, MainAd, urges agencies’ in-house teams to collaborate with external experts to further develop marketing efficiencies.
“Data scientists can be seen as expert interpreters who will get us all speaking the same language when it comes to data to drive more targeted campaigns and meaningful conversations not only between brands and their consumers but with their agencies, as well,” she explains.
Gender diversity should also be a factor to be considered when putting together data science teams say Boominathan, who manages an all-male team, as diversity unlocks innovation, not just in terms of gender, but also in wider considerations.
“Data science can make an impact on every industry, and therefore requires interactions with a wide variety of businesses and individuals. It is a field that values an assortment of skills, deep technical abilities, storytelling, visualisation, and constant learning,” she explains. “For gender diversity in particular, connecting the dots between the technical work and its social impact makes all of us more thoughtful, creative, and successful as a team.”
Boominathan adds that the growing awareness of data science is slowly attracting more women into the field, bringing a different perspective to the table, and we are seeing women easily rising into leadership positions. “For a woman to shine she has to find an environment that values her for what she brings to the table and celebrates what she’s good at. MainAd is one such company that values employees irrespective of gender.”
While their counterparts in the United States and Europe have already adopted data science fully in their marketing strategies, it is puzzling that agencies in APAC are only starting to embrace it fully now.
Mick Hollison, chief marketing officer at Cloudera explains that this is because of the dynamics that exist in the government entities in APAC, where there is a much bigger role in government and some of the GOs than in the US or Western Europe, and in China.
Hollison adds that it is also a historical thing as a lot of the technologies are invented in Silicon Valley and many of those technologies are widely used and adopted there. For example, the West coast of the US will adopt first, then the East coast, the Central and then the South.
“Even within the US, it is not like the country is one big thing and all of it happens simultaneously. It happens in a certain sequence, so it is the kind of the natural order of things,” he says. “However, there is some really impressive work that is taking place in APAC and as often is the case, there will be some technological skipping that will actually greatly benefit the region.”
“For example, the cell phone itself was invented in the US, but it got to GSM, which is a much faster way of delivery service in Europe. It then went wide spread adoption as a currency and the ability to pay for things in Asia. A much similar thing will happen here,” he adds.
While Nick Pan, regional head of planning and strategy at VML Southeast Asia and India observes that there have been lots of discussions by brands in APAC on setting up command centres to pipe in all possible data and allow different stakeholders across organisations to self-service and infer from data points, Sutton of AdParlor notes that as more and more tools become available and consumers’ demands evolve, enabling completely personalised marketing for consumers is an inherently iterative process and is a product of brands own business goals.
“This means all brands need to be taking steps to get more and more granular with their marketing messages. I think in Asia we’re moving more slowly on this journey than in other regions. I feel the key contributors to this slow pace of change are the lack of adoption of technology and the lack of support from the marketing ecosystem,” explains Sutton.
However, Hughes of DigitasLBi argues that it is not entirely true that APAC has not adopted data science quickly. He cites the example of China, which has been very quick to embrace artificial intelligence and continues to be a leader in that space.
“We saw some very early and sophisticated applications of AI on WeChat (chatbots, automation, etc) and with personalisation by the ecommerce giants,” he says. “But it is true that measurement and accountability for marketing across the region has lagged behind. There are a lot of contributing factors but certainly the comparatively smaller budgets in the region have played a role.”
As APAC has the highest number of adblocking figures, how crucial of a role data science will play in measuring and optimising campaigns for online and on mobile?
Liebmann of Exponential explains that while data science definitely plays a role in the adoption and usage of adblocking in APAC, it is just one contributing factor. “The biggest impact data science has on adblocking is when it is used in an unsophisticated way, for example, through the employment of re-targeting – this can very quickly shift the consumer journey into a state of annoyance.”
“What could have been a positive experience with a brand, then turns into a negative one that drives the consumer away from the brand and further towards adblocking as a solution,” she says.
Data science is crucial in reducing ad-blocking as it is a key component of enabling completely personalised marketing, Sutton of AdParlor stresses, adding that adblocking is a result of intrusive, high volume, high frequency and poorly targeted marketing messages.
“Natively placed advertising with appropriate messages based on consumers current behaviour and purchase intent is the future of marketing. Data science will enable the right ads to be delivered at the right time, in the right environment, and be optimised based on consumer responses,” he adds.
For Hughes of Data Science International, he opines that while it would be nice from an analysis standpoint and not privacy, to have pristine log files tracking every consumer behaviour, it is not reality and ad blockers are just one of many obstacles.
“There are still a lot of ways for data scientists to add value. The chief problem isn’t adblockers but rather deciding which of many potential analyses to focus on,” he says.