For many marketers, the first encounter with artificial intelligence has been surprisingly mundane.
Write a headline. Draft an email. Create a social post. Generate an image. Produce ten versions of an ad.
Those capabilities are useful. But the bigger opportunity isn’t simply using AI to create more marketing.
It’s using AI to make better decisions about who should receive marketing in the first place—and who shouldn’t.
Who should receive the offer? Which channel should we use? When should we communicate? Which message is most appropriate? And perhaps most importantly:
When should we do nothing at all?
That’s where AI starts becoming much more interesting.
From Generating Content to Making Decisions
AI adoption among marketers has accelerated rapidly. Salesforce’s 2026 State of Marketing research found that 75% of marketers are using AI, with personalization among its leading applications. Yet 84% still report running generic campaigns.
That disconnect tells us something important.
If AI enables an organization to create twice as many emails, ads, mail pieces and social posts—but they’re still going to the wrong people, at the wrong time, through the wrong channels—we’ve simply found a more efficient way to create waste.
Direct marketers should be asking a different question:
Can AI help us decide what to do next?
Increasingly, the answer is yes.
Maybe the Best Mailing Is the One You Don’t Send
Consider a traditional direct mail campaign.
An organization has 500,000 prospects. Historically, it might apply selection criteria, suppress obvious non-prospects and mail the remaining names.
Predictive modeling changes the equation.
Using previous response data, customer characteristics, engagement history, geography, digital behavior and other appropriate data, a model can help estimate an individual’s propensity to respond.
Perhaps 100,000 people have an extremely low likelihood of responding.
The smartest decision may be:
Don’t mail them.
Another group may respond particularly well when direct mail and digital advertising are combined. Use both. A high-value segment might warrant a more expensive mail package because its predicted lifetime value justifies the investment.
Suddenly AI isn’t helping us write the mail piece.
It’s helping determine whether there should be a mail piece at all.
USPS has highlighted this evolution, describing how AI and predictive analytics can help marketers identify promising audiences, personalize communications, optimize timing and learn which messages and creative approaches perform best.
And This Isn’t Just About Mail
The same principle applies across the entire direct marketing ecosystem.
Imagine a health plan communicating with members during Open Enrollment.
One member regularly responds to email and uses the member portal. Another rarely engages digitally but consistently responds to mail. A third frequently calls the contact center. Another has already completed the desired action.
Should all four receive the same communication sequence?
Probably not.
A decisioning model might recommend:
Member A: Email
Member B: Direct mail + email
Member C: Call-center outreach
Member D: Nothing—the member already acted
For another customer, the next best action might be SMS, paid social, CTV or some combination of channels.
The important shift is that the campaign doesn’t determine the customer journey. The customer’s data helps determine the campaign.
AI Can Help Answer More Than “Who?”
Audience selection is only the beginning.
AI and machine learning can help marketers answer:
Who? Who has the greatest propensity to respond?
What? Which offer, message or creative treatment is most likely to resonate?
Where? Which channel or combination of channels is most appropriate?
When? When is the customer most likely to respond?
How often? At what point does additional frequency stop helping?
And finally:
What next? Based on what the customer just did, what should happen now?
That last question is particularly powerful because customer behavior can continually inform the next action.
AI Is Moving Closer to the Customer
Recent developments suggest this shift is accelerating.
AI is moving from a content-production tool behind the scenes to becoming part of targeting, measurement, customer interaction and marketing decision-making itself.
Consumer expectations are changing as well. Adobe recently reported that three in five consumers surveyed said they would be more loyal to a brand offering a proactive AI experience that anticipates their needs. At the same time, 69% said clear disclosure that they’re interacting with an AI agent increases their confidence in a brand.
Better technology doesn’t eliminate the need for trust.
Better Data Still Beats Better AI
There is a catch to all of this.
AI is only as useful as the information and rules surrounding it.
Incomplete customer records, disconnected databases, poor response history and weak attribution won’t suddenly become valuable because an AI platform is analyzing them.
For experienced direct marketers, this should sound familiar.
We’ve been saying for decades that good marketing begins with good data.
AI doesn’t change that principle.
It amplifies it.
Human Judgment Still Matters
The mathematically optimal marketing decision isn’t always the appropriate business decision.
That’s particularly important in healthcare, insurance, financial services and other regulated industries.
Privacy, consent, HIPAA where applicable, fairness, brand standards, regulatory requirements and customer expectations all need to be considered alongside predicted response.
The goal isn’t autonomous marketing at any cost.
It’s better-informed marketing.
Less Marketing Could Actually Be Better Marketing
For years, technology has made it progressively easier to communicate with more people, through more channels, more frequently.
AI could accelerate that trend dramatically.
Or we could use it to do something much smarter.
Send fewer irrelevant emails. Serve fewer wasted impressions. Make fewer unnecessary calls. Mail fewer unlikely responders.
Spend more on the audiences and channels where the data suggests we can make a difference.
And stop communicating when the customer has already done what we wanted them to do.
That’s a very different vision for AI in marketing.
It’s not about replacing marketers.
It’s about giving experienced marketers another tool to answer the questions direct marketers have always tried to answer:
Who should we communicate with, what should we say, when should we say it—and what should we do next?
At Direct Choice, we believe that’s where AI becomes truly interesting.
Not when it helps us create more marketing.
When it helps us create better, smarter and more relevant marketing—and know when not to market at all.
Have Questions?
CONTACT US!
