Customers get it. In fact, we’ve hit a point where most people understand that automation probably played some role in creating, distributing or tracking the marketing message they received. The trick now is figuring out which aspects of marketing should be automated, where human judgment still belongs and whether all these tools are saving people time or just creating more systems to manage.

Selland tells companies to start with the work that slows teams down, versus the parts of the customer relationship that still need a human touch. His short list:

1. Test out high-volume, rules-based tasks.

Lead scoring, follow-up sequencing and campaign management are good testing grounds because they require consistency, speed and repeatable rules. Data cleanup can also be a good place to start, especially for marketing teams dealing with customer information spread across too many point solutions.

“Those tasks don’t necessarily require judgment; they’re more about consistency and in some cases, speed,” Selland says. “Anything that’s high-volume and rule-based, that’s a good place to use AI.”

2. Harness LLMs for specific jobs.

Companies can get more from large language models (LLMs) by putting guardrails around them and tailoring the tools for specific marketing needs. Selland points to Claude as one example. He’s been experimenting with newer design-oriented AI tools for his own business, using them to help with branding and visual development without having to learn the fine points of Adobe or Figma.

Used the right way, AI can help marketers cut down on repetitive tasks, clean up messy systems and get more from the tools they already have. It still takes people to decide where those tools belong, what should be left alone and when the outreach has crossed the line from useful to too much. The person on the receiving end has tools too, and if the message feels excessive, irrelevant or too automated, they can shut it down just as fast. “It’s okay to do more with less,” Selland says. “But just don’t overdo it.”

3. Whittle down the marketing stack.

As automated marketing tools have proliferated, so have the number of software subscriptions that teams have to pay for, manage and justify. “There’s a lot of excitement about being able to do more with less, and more pressure to be both efficient and effective,” he says. “Right now, the CFO is asking the CMO: Why do we need 10 SaaS products? Can we just do it with two?”

Selland recommends picking one or two strategic vendors to work with, then getting more out of those relationships instead of trying to manage, use and train on a group of disconnected marketing tools. That gives the team fewer systems to learn, fewer contracts to defend and fewer places for customer data, campaign activity and performance results to get buried. 

Use customer success feedback to guide automation

CMOs using automation and AI to speed up marketing also need a direct line into what customers are saying once those campaigns, messages and automated touches reach them. Sales departments may control the “new prospect” conversation, but customer support and customer success teams hear from the people who already bought, used the product or service and formed an opinion about the brand. 

That feedback can help CMOs understand where automation helps, where it gets in the way and what customers still value about the company. So before adding another tool or launching a new sequence, Selland recommends spending more time with the customer success team, which is “always going to be the best source of information about what’s most valuable about your brand from the customer’s perspective.”