Computers and artificial intelligence have come along at an exponential rate over the past few decades, from being regarded as oversized adding machines to the point where they have played integral roles in some legitimately creative endeavours.
Emarsys CMO, Allen Nance, might believe we’re a fair way off from general purposed artificial intelligence, but he’s convinced that when it hits, it’s going to unleash a renaissance in creativity.
“If we’re being truthful, digital marketing isn’t very creative today,” Nance told CMO during a recent visit to Sydney. “But through machine learning, the next decade could see the rise of a Mad Men version of digital marketing, where creativity goes through a renaissance and we start to see really interesting, emotional videos and content.”
Nance joined Emarsys as CMO last year to help the marketing technology company realise its ambition of becoming a US$150 million software business.
Emarsys provides an omni-channel marketing technology platform aimed at B2C organisations. The 15-year old Austrian-based vendor has nearly 2000 clients globally, manages more than 250,000 personalised campaigns per month and has annual revenues of US$60 million. Clients include Cosabella, Catch, Styletread, Runtastic, Toys R Us, Sky and Clink Hotels.
Telling the story of an omni-channel marketing platform
An engineer by degree and computer scientist and developer by trade, Nance was responsible for coding one of the original email marketing platforms, WhatCounts, delivering flight confirmations to passengers. After selling that business, he spent several years as an early stage technology investor, supporting 18 different software companies before stumbling upon Emarsys.
Nance said he was initially attracted to Emarsys because of the vision and entrepreneurialism of its Israeli-born founder, Hagai Hartman. After spending 3-4 months getting to know the technology, and visiting the R&D facility in Budapest, he became convinced Emarsys had built one of the few single platform engines that could execute true omni-channel marketing.
The third thing was the need Emarsys had for an executive leader to corral the 80-strong marketing team globally and help better tell the vendor’s story.
“Our MD thinks of marketing as a product data sheet, not as a story. He can articulate the product does these six things, which is good, because a company can’t build a brand on lying, but I’m the opposite: I’m much more of a storyteller,” Nance said. “I don’t think marketers buy data sheets, they buy aspiration, stories and visions. They buy into a place that’s going to help them get there. The product isn’t quite there but it’s the closest single-build product I’ve seen.
“There are a lot of companies building platforms in this space but the reality is a lot of them are an amalgamation of four or five acquisitions that have been made.”
How artificial intelligence will change marketing
One of the key areas of investment in marketing technology right now is artificial intelligence (AI). As vendors such as IBM, Adobe, Google, Salesforce and Emarsys look to build AI and machine learning into their offerings, significant questions are being raised as to how these new technical capabilities will change both the way marketers execute activities, as well as the job of marketing itself.
“Humans cannot drag-and-drop their way to personalisation. It doesn’t scale. Everyone knows that,” Nance said. “The real opportunity I see for AI and machine learning is to actually scale personalisation beyond the human capacity.”
Nance is quick to point out machine learning is not new, nor is it necessarily the competitive advantage in and of itself. “Every single person who has ever opened up Google and misspelt a word to get the correct spelling in the search string has used AI,” he said.
“The way I describe it is to look at Google-owned DeepMind, the foremost experts working on general purpose AI, and an actual machine that can learn and predict the next best action to take. Their learning algorithm is free. Why is that? Because an algorithm without good data is useless.
“Marketers still have to solve the problem of having crap data. These algorithms are not going to work if they don’t have great data.”
Nance also sees a problem with the way many vendors are approaching AI – namely, as a module or bolt-on capability to an existing martech or enterprise stack.
“AI will struggle in many places… because it can’t be separated from the data, or be something you can bolt on to the side of your products,” he claimed. “The whole idea of machine learning is I can take an algorithm, apply it to data and learn something from it. The primary way I do that is through pattern recognition – it’s not a new technique.
“When I look at AI and its implications on marketing, it’s a simple concept people are making too hard. And that is human-driven personalisation at scale.”
But to do this, AI must be vertically integrated into a platform, accessing the data foundation, intelligence layer, execution layer and deployment engines.
“If those are the layers of our marketing cake, the reality is AI can’t now be a module, it has to vertically integrated with all of that – it needs the data, the intelligence, the segmentation, then access to deployment engines in order to execute a campaign,” Nance explained. “The best companies investing into AI are integrating it as a technical capability.”
Why marketing automation isn’t working
One reason why AI is gaining so many column inches today is that marketing automation has not delivered on its promise, Nance claimed. “It made things better for the marketer – before, marketers were batch and blasting, and this is an incremental improvement on it. But if the promise was personalisation, we’re not there yet,” he said.
“Everyone is chasing personalisation, but I think the mistake we made eight years ago is that we took marketing automation and personalisation and slammed them together. That became the promise when actually they didn’t have anything to do with each other.”
The second reason for AI’s rise is that the technology has matured to a point where it can be productised, Nance said.
Get your data and content right
“Then as marketers do, we seized on it. But if we’re not careful, we’re going to overpromise on this thing,” he warned.
So what do marketers need to do to avoid misreading AI’s impact and role? For Nance, the first thing is getting serious about data management.
“Serious doesn’t mean going to buy a new product, it means re-engineering our processes, having data quality processes, and understanding every touchpoint I have with the customer to ensure I’m collecting data,” he explained. “Data is not a technology problem, it’s a process/strategy problem.”
The second thing marketers must do if they’re serious about personalisation is to understand that it’s a content problem. “If I’m going to deliver four different pieces of content to meet those personalisation needs, do I have those four pieces of content?” Nance asked. “When I ask marketers that question, about 90 per cent say no.
“I’d love to see marketers spend more time on the two ends of the spectrum: How good am I at my data, and how good am I at my content. Apply a bad algorithm to bad data and one image, and it isn’t going to work.”
For Nance, more marketers also need to start separating the concept of content from the concept of a campaign.
“Where marketers still start is with a campaign, then they pick a channel, then they develop content – it’s in that order,” he continued. “In the space we deal with most, which is ecommerce and retail, I’ve tried to advocate taking the word ‘merchandising’ and replacing it with ‘content’. The content emulates from the point I pick a fabric, colour, or theme.
“In the operating of your business, you’re developing content, you’re just not functionally doing it.”
The marketing skillset
There’s no doubt all of this still represents a revolution of the role of the marketer. But instead of technical smarts, Nance believes key attributes for long-term success will be creativity and strategy.
“I never meet a marketer who went to university to study marketing so they can drag-and-drop in Marketo,” he pointed out. “If you got into marketing, you probably thought it was because it was sexy, creative. As it’s become more technical and digital, there are bunch of people doing something they hate.”
Over the next 36 months, as machine learning picks up the ability to do segmentation, choose the appropriate channel, time of day and format of messaging, marketers are going to gain the opportunity to go back to strategy, content and creative, Nance claimed.
Emarsys is looking to help make this shift in technology capability for marketers, and its teams have set themselves the task of breaking down and automating a traditional win back campaign for a retailer over the next 36 months using AI capability.
“There are 28 decisions and 10 pieces of content marketers have to deal with in this campaign. We are taking each decision – which humans are making right now – and asking what coding is required for the machine to make each decision,” Nance said. “We are deconstructing this into a series of machine learning algorithms and AI predictions. Our goal is to take 28 decisions and turn that into two decisions.
“We’ve come up with a term ‘artificial intelligence marketing tuning’, which is a series of sliders and that’s how users will give the machine the strategy. Then you’ll load content into the content centre and manage the creative.”
By partnering with intelligent machines, marketers will be able to spend most of their time on strategy, content and creative.
“Multi-channel campaign deployment using AI is going to become commoditised and once we get there, who is going to win? The people with the best data and the smartest people who develop the best strategy, content and creative,” Nance said.
“And that’s a battle CMOs know they can win. They’re back to what they’re good at.”