Buyer beware: Marketing automation isn’t really automated

The latest Annuitas B2B Enterprise survey of over 100 B2B enterprise marketers in organizations with annual revenues that exceed $250M revealed that only 2.8% of respondents believed demand generation campaigns achieve their goals.

The responses laid the blame at a variety of doors: sales-marketing misalignment, measuring the wrong metrics, lack of quality leads, and so on. Undergirding all of these responses was frustration in the technology powering campaigns - marketing automation.

‘Set it and forget it’: The false promise of marketing automation

Marketing automation has been around for years and in that time the pitch has remained largely unchanged: marketing automation will take previously manual tasks and, well, automate them. This has led to the misperception that marketers can just ‘set it and forget it’ - underestimating both the amount of work required by marketers to make this technology work effectively and the necessary maintenance required to optimise such efforts.

Marketing automation only works when there is a process in place. Take, for example, the act of sending an email campaign: Content strategy and creative direction is needed to devise what goes into the email; list and database management is required to decide who receives it; a campaign manager or email channel manager decides when it goes out; a demand generation exec has to report on the results; and a data strategist must pore over the campaign results to inform strategy for the next email campaign.

None of these decisions or processes are suddenly automated away by marketing automation. Marketing automation doesn’t automate your marketing so much as streamline and scale your current processes - it’s a workflow tool, not an automation tool.

Buyers don’t care about your (automation) rules

They say that rules are meant to be broken and this is no less the case when it comes to marketing automation. The backbone of marketing automation is preset logic (“If this X happens then do Y”, “if X does not happen, then do Z”) or rules that are used to architect marketing campaigns and trigger communications.

Many marketing automation platforms still base their design and application on the purchasing funnel developed by E. St. Elmo Lewis in 1898. This model has its merits but it ignore the complexity of the modern day B2B buyer whose journey has changed drastically since the invention of the purchase funnel. Consider the Internet, online reviews and social media and their impact on how we made purchasing decisions today. In particular, we know that in a ZMOT world, content increasingly influences an individual’s process and not necessarily the brand’s content alone. Buyers don’t follow your campaign process, they follow their own!

Marketing automation doesn’t give context

There’s a bigger issue at stake: marketing automation is only as good as the customer data that is used. The stock customer data in a marketing automation database includes contact details, firmographic details, known purchase history and a lead score based on a cumulative amount of scored ‘interactions’. 

Marketing automation ignores that buyers are continually evolving in their interests, needs and motives. Whilst marketing automation will recognise if a person is stuck in a particular campaign phase – perhaps they have recently opened a second brand newsletter but did not click on the CTA button at the bottom – your marketing automation will not be able to tell you that the person is considering changed jobs or considering other solutions; something that could be revealed by Content Intelligence.

Marketing automation needs to get better at understanding and logging buyer context. Not just because it will allow marketers to align their messaging with buyer needs, but also because it has knock-on effects when a lead is handed over to a sales rep. The lead may well carry a name, email address, job title and score, but that is of limited use to a sales rep that wants to have a relevant and engaging conversation with their prospect.

So, so-called marketing automation stands guilty of several charges: the need for constant human intervention, beholden to rules-based logic that is unfit for the modern purchase journey and, finally, unable to record and accurately identify emerging buyer interests.

Machine learning for marketing automation

Machine learning refers to a kind of computational data analysis whereby algorithms ‘learn’ from new information and quickly decide what the next best action is for an optimal outcome.

Machine learning is well-suited to the new Big Content environment where CMOs face complex buyer journeys, constantly evolving user profiles and myriad pieces of content that need to be categorised and structured before being served across multiple-platforms.

It really is the only way to uncover trends manual human analysis would likely overlook. In a marketing context this could, for example, mean identifying and displaying the most relevant content topics or themes that are resonating with your audience at any given time.

Rather than relying on restrictive rules-based logic, machine learning adapts to the unique signals and interactions of each buyer and automatically decides the next-best-content to send to them. Similarly, an algorithm can learn each individuals’ interests by analysing their unique content consumption. This could be returned to CRM software for customer insight that goes well beyond most ‘Voice of the Customer’ programmes.

For those 97% of B2B marketing execs who felt their demand generation was not successful more marketing automation is not the answer. Any improvements from hiring better staff, proactively aligning marketing and sales around the customer, and improving lead management will be welcome, but incremental at best. If marketing automation wants to be truly automated - and start delivering better demand generation campaigns - we need to start talking about machine learning.  

Munya Hoto is VP of demand generation at idio.