Some may conflate segmentation or product recommendations with personalisation. For others, personalisation means nothing more than ‘[first name] [last name]’ in comms.
However, martech integration and the application of machine learning is now enabling more sophisticated personalisation that truly deserves the name. As this AI tech becomes easier and cheaper for marketers to adopt, marketing roles are slowly being redefined.
All this is easy to observe in the transformation of email service providers into ‘marketing platforms’, ‘personalisation platforms’ and other soubriquets. Though vendor hype may run a pace ahead of what’s happening in the market, the future does seem close.
The idea and reality of personalisation is what I wanted to discuss with Raj Balasundaram, VP Solutions and Strategic Services at Emarsys, a B2C marketing automation platform.
Every marketer has to answer four questions
I began by asking ‘What is personalisation?’
“Every marketer has to answer these four questions,” Balasundaram replied, “Who? What? When? How?”
“That’s fundamentally what personalisation does. Who is the customer? What am I going to say to them? When, or in what context? And how am I going to deliver that message?
“The four questions,” he continued, “need to be answered at an individual level, and they need to be answered every time we contact the person and without thinking about what channel we’re going to send to.”
It’s this concept of lots of individual decisions being made, each considering some aspect of content, time and channel that makes this personalisation different.
Balasundaram simplifies it for me: “The machines don’t segment, they don’t personalise, all they think of is an event. So, ‘here’s Ben, what do I need to do with him?’ It’s a singular transaction, rather than putting a list together or using smart content blocks for example.”
Essentially, this is marketing automation but with many more variables. Rather than designing a handful of pathways which a consumer might be funnelled down (e.g. welcome campaigns, loyalty campaigns etc.), the technology uses statistical analysis of the information that the marketer has about the individual to decide what the best action or option is in any instance.
“It’s not the channel that surprises [customers], it’s the content…”
Balasundaram remarks that this tech is effectively bringing an end to siloed marketers. He says that “Whereas currently [marketers] have already decided what they’re going to do – ‘I want to send an email, I’ve already decided Ben is in this particular group, and I’ve already decided what the content will be’ – this is not what personalisation is about.”
What Balasundaram is referring to is channel agnosticism. And while some marketers may think this ignores the fundamental difference between media channels and content formats, Balasundaram is also advocating for a return to a more strategic way of thinking.
“It’s not the channel that surprises [customers], it’s the content that surprises them,” he says. Though he does point out that millennials are more likely to be delighted by personalised direct mail simply because they may never have received it before.
“The content [or message] should be created well before we decide to go down an email route,” he continues, “and this takes away the need to do segmentation – I already know what to say to Ben, and I’m finding the right moment to say what I want to say, and that is vastly different to the way marketers work. It’s a different way of thinking.”
“If you start segmenting people, you’re really not personalising…”
Balasundaram can sound quite dogmatic – “[Email marketers] have been doing the same thing over and over again, and it clearly doesn’t work, people know it’s a mass email. Even the common consumer knows it’s mass emailing,” he says. But he is also realistic and recognises that common practices in marketing are influenced by the technology available. Take this soundbite for example:
“If you start segmenting people, you’re really not personalising. But I don’t think there’s a difference between personalisation and segmentation, they are one and the same – the reason we did each is purely down to the level of tech we have or the limitations we have. Now the tech is taken care of, should we really go back to segmentation?”
This was the part of our discussion where we got to the crux of the matter and Balasundaram’s most illuminating point.
“So far,” he says “marketers have been concentrating on the operational part because to get a campaign out the door, it will take them two or three weeks to arrange the data, all the coding, segmentation – which is internally focused, operationally focused. And they actually end up not concentrating on the most important thing, the creative part.
“[This] was not the marketers fault, the tech didn’t help them out, but now the whole work paradigm will change simply because of the fact all we expect marketers to do is write content for their end consumers. The tech forces marketers to think about consumer perspective every step of the way. When an email goes out and the marketer looks at it and says ‘yeah, I know it’s not perfect, but this is the best I can do’ – that will change, because marketers have fewer excuses now. The tech has caught up to a point where you can go individual to individual.
As an addendum, Balasundaram says “You can even generate the content using AI”, referring to tech such as subject line optimisation which is rapidly being adopted by big brands that send do a lot of marketing messaging.
“A pure email marketer probably won’t exist in the next five years.”
The shift of mindset to customer-centric campaigns, away from operational-centric campaigns is what Balasundaram describes as “taking the [channel] silo away, putting everything into a common pool and finding patterns in it.” From a tactical point of view, this could entail using push notifications for users that don’t open emails, or search retargeting for those that unsubscribed from email, perhaps with an incentive to return (such as free delivery).
Typically, Balasundaram tells me, Emarsys will work with an inactive part of a client’s customer database when that client first trials their machine learning tech. He says they may look at “churning customers, or customers about to leave or not responding…then apply AI personalisation techniquesand…it usually takes about 6-8 weeks for the algorithms to learn a bit more about the customers but then they’ll eventually see the results.”
When I ask what this means for the marketer in the long run, Balasundaram is punchy. He says “A pure email marketer probably won’t exist in the next five years. They need to think about email marketing in terms of a bigger business strategy. If they’re going to be pure email marketers, it will be difficult – if you don’t see the customer as part of the bigger picture, it’s never going to work.”
He continues, “Marketers will have more time to think about business strategy and tactics, and the components required in creating the content. They can spend more time… creating rather than deploying. Instead of thinking about improving clickthrough rate, they can be reporting on revenue. [It’s about] revenue over operations.”
This is a familiar yarn, but marketers do seem to be getting there.
I can’t help but wonder if the best preparation marketers can do is get right back to basics and try to forget about the technology altogether.
Thanks for reading. N.B. Econsultancy runs a variety of email marketing and CRM training courses. Get in touch for more detail.
Article by channel:
Everything you need to know about Digital Transformation
The best articles, news and events direct to your inbox
Read more articles tagged: