A lot has been said about striking the balance between the art and science of marketing. For martech stacks, there is a similar need to balance. On the one side, you have automation-powered efficiency, which delivers speed, scale and avoids human error and limitations through tools such as email marketing, social media publishing, web testing and more. On the other side lies intelligence-powered effectiveness, providing iterative, intelligent, on-the-go improvements in outcomes through tools such as decisioning engines, analytics tools and personalization solutions.
So should marketing leaders build stacks for efficiency or effectiveness? Is there a rule of thumb to make choices that drive optimal business outcomes and stack ROI?
Marketing Efficiency vs. Marketing Effectiveness?
Almost all the executives we spoke to recommended designing and building the martech stack for effectiveness, not efficiency. Or as Scott Brinker of Chiefmartec and VP platform ecosystem at Hubspot put it, “If the marketing isn’t effective, it doesn’t really matter how efficiently it’s failing.”
When does automation-led efficiency matter? “It’s a no-brainer that automation works in repetitive and highly predictable tasks that need to be managed at scale (without scale, the investment in automation tools doesn’t pay off). Testing applications and websites is one such use case, in the Martech domain,” said Kshitij Mulay, a Singapore-based CIO of a global CPG brand.
Brinker added that the more stable and repeatable the process being automated is, the easier — and safer — it is to automate it. “One of the reasons I’m a fan of no-code tools for automation is that it makes it relatively simple for marketers to make iterative changes with minimal overhead.” Vivek Chandrashekhar, director of growth marketing at ComplianceQuest and a practitioner who spends heavily on digital, highlighted the need to focus on the efficiency of tactics before measuring effectiveness and ROI of campaigns. “Take programmatic advertising as an example. Unless the programmatic tactics such as keyword choices, CRO, CPC optimization, and even page-level technical KPIs are set up and running efficiently down to the smallest detail, the room for lead leakages is significant. Tactics must be efficiently set up and running before outcome measurements can be considered reliable or effective.”
Related Article: How COVID-19 Changed Marketing
Making Effectiveness Work in the Martech Stack
Unlike automation, driving effectiveness is not quite as straightforward. The problem is of course that neither efficiency nor the presence of intelligent, so-called AI powered tools is a guarantee of effective marketing outcomes. Intelligence is an outcome of human and machine working together, and needs clarity on the problem being solved and appropriate in-house skills, among other factors.
In a conversation on the popular Slack channel, Analytics for Marketers, Christopher Penn said the choice (of intelligent tools) depends on your goals and business requirements. “What’s the bigger problem, poor results or overstretched humans?” Unfortunately, we often see AI being invested in and deployed for the sake of it. If that were not the case, said Penn, we wouldn’t, as a profession, think 0.2% ad performance is great: “In what other reality is failing 99.8% of the time considered good performance?” Perhaps because “Like any tool, your skill with it determines your outcomes,” as Katie Robbert wrote in the same thread.
Ongoing effectiveness also needs a more adaptable martech stack architecture, said Brinker. “Marketing and technology both change so rapidly these days that it’s super valuable to have a martech stack that can evolve quickly and easily to address new opportunities.”
Effectiveness also needs laser-focus on the right goals. All else is wasted effort, said Anand Thaker, executive growth advisor and martech industry insider. “I’ve witnessed millions worth of transformation efforts wasted because the focus was on savings over enabling customer-centric growth. Speed of execution matters, but only in the right direction.”
Related Article: Does Your Martech Stack Inspire Joy?
Avoiding the Effective vs. Efficiency Sinkholes
Ultimately, no matter how good the tools, both efficiency and effectiveness investments can under-deliver if marketers don’t set themselves up for success. So, what are the factors to consider? Readiness is a success factor for either, said Brinker. Automation products work best not just when there is clarity around processes, but the ability to integrate with the technology in your stack. Similarly, he added, intelligence products work best when good, clean and relatively complete data is available to work with.
Rather than falling for the hype around AI, Chandrashekhar focuses on determining fit for specific business context and operational constraints. “Context includes buyer personas, channel preferences, and sales and marketing goals, and constraints are budgetary allocations and limited resources. Understanding these two at a granular level should lead the decision-maker to prioritize what needs automation or intelligence.”
Companies also make poor tech choices when they lack a data collaboration and growth culture, added Thaker. “Teams must be allowed a safe space to experiment, fail, test the limits of their capability, and be transparent about the outcomes. Without that, they cannot see the opportunities beyond the box they allow themselves to be placed in.” Equally damaging can be misaligned incentives, said Brad Bauer, senior manager of digital marketing at ecommerce marketplace IAA. “Marketers that are measured on activity often want to automate so they can continue to ‘dial up the activity,’ but that often creates outcomes that are unproductive or even downright damaging to the brand.” Think “Dear [Name]” emails dropping into your inbox by the dozens every day.
Related Article: How Many Marketing Emails Is Too Many Marketing Emails?
It’s All About That Balance
What it comes down to is designing and building a martech stack for effectiveness, while leveraging automation tools to scale proven processes. But remember, deploying AI tools does not automatically translate into intelligence or effectiveness. Marketers should first think about what kind of intelligence a particular process needs. Once that logic is addressed, the right AI tools, used by appropriately skilled marketers, can be effective in continuous improvement at scale, joining many more dots far faster than any human can hope to do.