Welcome to B2B AI – boring, but effective

AI isn’t all about voice interfaces and self-driving cars, it can help B2B marketers address more immediate concerns like improving data analysis and decision making.

I have written a number of articles about machine learning recently for Marketing Week and Econsultancy. The current ‘AI summer’ is of course generating a lot of column inches in the mainstream media, too.

This interest is fuelled by consumer tech – if you were to play word association, ‘AI’ might be followed by ‘Alexa’ or ‘self-driving cars’ – as well as bullish claims about the potential of deep learning in medicine.

As marketers we should be much more prosaic in our thinking. That’s not to say that voice interfaces aren’t going to have a big impact on your brand and the discovery of your content – you will have to consider what your brand’s persona will sound like, and how you will answer user searches directly.

But the more immediate and broader concern is how data analysis and decision making can be improved. This is what consultant head of best practice reports at Econsultancy, Steffan Aquarone, describes as ‘automation-plus’.

B2B marketing is a great crucible for machine learning and automation-plus. Though B2B marketers may call on many of the same machine-learning technologies as B2C marketers, it’s the application in lead generation that is most revealing. Predicting which contacts are most likely to convert throughout the consideration and purchase journey, and serving them the content and comms best suited to tickle them along the buying funnel, is a process that is quickly tending towards automation-plus.

READ MORE: Ben Davis – AI is much hyped but often misunderstood

Making the most of machine learning

Account-based marketing (ABM) – defining your best prospects and marketing to them with personalised messaging and content, rather than casting your net wide – has seen vendors rush to trumpet their machine learning capability.

Software from companies such as Nudge and Demandbase will sort through lots of publicly available data from news, regulatory filings and the social web, then use natural language processing and relationship mapping to identify potential clients and their decision-makers. Personalised comms can be created from this information and passed to the salesperson ahead of the first outbound contact. What’s more, the software can determine who in your sales organisation is best placed to make that first contact.

Innovative customer relationship management (CRM) systems such as Salesforce will use ‘predictive lead scoring’ to try to understand which accounts are warmer than others, and where the salesperson should spend their time. Such scoring may draw from data sources such as web browsing behaviour and content downloads, as well as email opens. This is familiar territory for marketing automation, but with a layer of ‘intelligence’ that can help to remove guesswork.

Much like in the world of advertising and ecommerce, CRM and ABM software can automate segmentation of website content, target lookalike audiences and continuously optimise for the best chance of success with a particular customer profile. Optimisation may also involve some level of attribution, pointing the marketer towards more efficient marketing. Platforms such as People.AI can also monitor which accounts salespeople are spending time on, and flag up at-risk clients.

If there’s one thing we can agree on, it is that the human brain is too valuable to be wasted on data entry.

This is all quite boring stuff to some. But the same people may be bored by conversion optimisation in ecommerce, a technique that relies on similar technology and which has been a big part of Amazon’s rise.

Aquarone’s pragmatic definition of automation-plus, and its obverse – the bombastic claims of B2B software vendors – are signs that data-backed decision making is becoming more sophisticated in marketing.

We shouldn’t get complacent and see these solutions as some kind of black magic. They are often rule-based and similar to the algorithms that have been optimising advertising impressions for years, rather than cutting-edge unsupervised machine learning of the like used in visual recognition. However, these solutions give us a glimpse of a future when marketers can devote more of their time to insight.

In the past, B2B salespeople and marketers have spent all their time blasting out comms, calling customers, doing lots of relationship building and manual CRM work. More recently, marketing automation has cut out some of this leg work, but the business of creating and keeping on top of automation pathways and content represents simply another type of leg work.

Now, though, we are starting to see machine learning used to allow marketers and salespeople to prioritise, and that is surely a salve to every practitioner out there who is still snowed under and wondering when the robots will finally arrive and give them some time to think. If there’s one thing we can agree on, it is that the human brain is too valuable to be wasted on data entry.

READ MORE: Are robots after your job?

Ben Davis is editor at Marketing Week’s sister title Econsultancy.

Don’t miss Marketing Week’s AI and technology conference Supercharged, taking place on 1 May in London, where you can learn how AI, machine learning and emerging technology are powering marketing. Register now.



Marketoonist on artificial intelligence

Tom Fishburne

Tom Fishburne is founder of Marketoon Studios. Follow his work at marketoonist.com or on Twitter @tomfishburne See more of the Marketoonist here Tom Fishburne will be speaking at the Festival of Marketing, which is taking place on 4 and 5 October at Tobacco Dock. To find out more information, including how to book tickets, visit […]