Robert Sawatzky
Feb 5, 2020

IBM CMO: With great technology comes great responsibility

“In many ways, Watson was the brand that introduced AI to society,” Michelle Peluso says.

Michelle Peluso
Michelle Peluso

We know marketers have more technology and tools at their disposal than ever before, but some are blessed with a far greater bounty than others. Imagine being the CMO of a company that brought artificial intelligence mainstream, pioneered cloud computing, became a global consultancy leader on data management, has a pipeline of world-renowed computer engineers and data scientists from the world's top tech institutes and owns legendary supercomputer brands in Deep Blue and Watson.

“The good news [for me] is I get to work at this incredibly storied brand," IBM’s chief marketer Michelle Peluso tells Campaign Asia-Pacific. "What we have offered from a product perspective has changed many times over our past 109 years but what hasn’t changed is our values. So we offer a clear proposition to the world: That we roll up our sleeves each and every day with our clients to bring the most innovative technology and rich industry expertise to bear to help solve the world’s hardest problems.”

This may sound rehearsed, but the get-to-work mentality isn't. Peluso's sleeves are definitely rolled-up as she squeezes our interview in between a "packed" day of meetings and events in Singapore atop a downtown office tower, fresh-off a panel discussion with brand clients around data trust and transparency inside their 'IBM Studios' facility. And the problems are indeed large. Finding workable solutions around sustainability, smarter government and digital transformation brings its own kind of pressure, putting the seemingly smaller challenges around marketing into perspective. 

But in her marketing work, Peluso also has some distinct advantages. Her teams get AI-driven alerts throughout the day around what's working with their campaigns, what’s not working, where they may have low engagement rates and where their audience targeting might be off. “We’re using AI to help the marketer do her best work,” she says, moving on to explain how Watson Ads are steering advertising towards conversational and reciprocal interactions between brands and consumers through all devices. New tools, like IBM's recently announced Advertising Accelerator, help marketers predict what creative elements best resonate with their audiences. 

But with such an embarrassment of riches in technology at her disposal, the bar is set high for a CMO like Peluso to deliver on her own work, to show the technology can be applied and implemented properly much to the envy of marketing clients. 

So how does she deal with that pressure and responsibility?

“I hope we always keep the bar high. It’s a good thing.” Peluso asserts before continuing, “I actually think in some ways, the most mundane things matter most in that question.”

The more advanced the technology and data capabilities are, Peluso explains, the more important it is for marketing teams like IBM's to get the basics right.  Included on this list for her is data governance, an agile workforce, diverse teams, bringing more in-house, being selective with partners and protecting her own brands.

Data governance

While admitting data standards and governance is not a scintillating topic, it’s something Peluso says she’s “incredibly passionate” about. To do their best work, data scientists need to be able to access and distribute data broadly. But in this post-GDPR world, it’s critical to be done securely and consensually. 

This is why much of Peluso’s work is invested with privacy, governance and security teams and why she hosts events like client forum on ‘Trust and Transparency’ in Singapore, sharing findings from their global C-suite report on data siloes and sharing. 

An agile workforce

With the right data governance in place, IBM’s zealous commit agile workflow is another key basic, Peluso contends, since “small cross functional teams that have access to democratised data can make better decisions on behalf of the customer.”  Here, she cites a variety of clients from banking to airlines, where customers are crossing multiple silos from online to offline, with external conditions like weather and markets that can affect service outcomes. With all that data residing in different parts of the organisation, it’s imperative that customer-centric agile teams be able to collect it, synthesise it and act on it in a way that elevates service.

Asking the right questions

“You can drown in the data,” Peluso admits. “Again, it sounds very mundane, but the teams I see do that the best work both on my own marketing team but also outside are the teams asking the best questions. There are other teams who just wade and wade in the data and they get lost.”

The questions that give focus, Peluso says, can include business basics like: How to increase loyalty for a certain segment? And what can that business result be? It’s also asking how technology like AI can systematically be applied to better understand consumers, interact with them and in the same way help a marketer with each of her tasks throughout the day.

Diverse and inclusive teams to root out bias

If you want teams asking well-rounded questions and technology applied in a fair way to everyone, “inclusion and diversity matters more than ever,” Peluso contends. “It's so important how we train AI, how we usher in new technologies that we do so with diversity and inclusion at the forefront, otherwise we don’t like the world we end up inheriting.”

Here she cites an Amazon's 2018 decision to scrap its AI tool for screening resumes that was biased towards male candidates.  She also points to evidence to showing AI doesn’t recognise brown and black faces as well as white ones and what IBM is doing to change that

In January 2019, IBM Research released ‘Diversity in Faces’ dataset to advance study of fairness in facial recognition systems

“Past experience can lead algorithms to sort of be trained in a way and we don't like the consequences. And It's both our own bias, but it's also the data itself,” says Peluso. “So how do we usher in this new era of AI, grounded in principles...with really diverse teams that can be on the watch.”

Peluso contends identifying AI bias will quickly become essential if not mandatory for regulated industries such as banking.

“You can't make credit risk decisions if you can't trace the decision back through the algorithms.  It’s not enough to say: ‘Our AI algorithms said I should lend to Howie, and not to the three of us, unless we know for sure that wasn't because the AI system exhibited a bias towards women. So traceability and transparency will be really critical.

Keeping it close to home

Since IBM wants teams that are agile, diverse and unbiased, it makes sense that she would want to bring more marketing in-house to have more control over the makeup of her teams. 

“[In the] big picture we probably do more in-house today than we did a few years ago, but we’ll always have critical partners. The second thing is we work with [fewer] agencies today than we did three years ago.”

These include George P. Johnson on events and select WPP creative and media agencies depending on the nature of the work.  The WPP agencies have had IBM agile training and now work off the same data sources as IBM. 

“If you really believe in the principles of data sharing and in agile, you just can’t do that with everybody,” Peluso says. “So we’ve really streamlined the portfolio of agencies, doubled-down on the ones that matter and changed the way we operate with each of them.”

Protecting your best brands

If being selective about agency partners is one way to safeguard IBM’s technology and reputation, then perhaps even more care needs to be placed on its technology partners.

Companies of all stripes are constantly trying to get their hands on an AI like Watson for myriad applications. In the marketing world alone in recent years, we've see IBM Watson partnerships with Salesforce on customer service, MediaMath on media optimisation, The Weather Network on dynamic creative advertising among the various pacts with local agencies and consultancies on using their tools and technologies. 

IBM's Watson wins US trivia game show 'Jeopardy' in 2011


“The Watson brand was built with a lot of care and precision so we’ve got to be very thoughtful about where we partner,” Peluso says. “In many ways, Watson was the brand that ushered the concept of AI into society. Even in very simple ways – like with Jeopardy (the trivia game show where Watson soundly defeated top human contestants in 2011) – it introduced artificial intelligence in a way that humanised it and created a discussion around AI.”

That kind of experience in humanising technology to speak to consumers is invaluable to marketers, and while Peluso feels privileged to work with it every day, she says all marketers are now working with all sorts of technologies in similar ways, at the intersection of customer experience, data, art and creativity, which is a great spot to be in.

“What I love about marketing is it's become a growth job.”

Related Articles

Just Published

1 hour ago

Glenfiddich asks 'where next' with story of stag in ...

Watch the film conceptualised by Space (UK) here

2 hours ago

How to successfully run a virtual event (and ...

The analogue era of face-to-face events may be coming to an end, and this means that marketers need to ramp up the creativity and engagement virtually.

2 hours ago

Snap revenue climbs, but monetising new users ...

Social media platform's revenue rose 52% for the third quarter, daily active users climbed 18% and overall per user revenue jumped 28%.

3 hours ago

Hakuhodo acquires stake in UK digital transformation...

The Japanese holding company has the option to increase its 25% equity stake in Public Digital to a majority interest in the future.