You Do Not Need an AI Strategy to Start Using AI

Aug 11, 2026

The pressure to “have an AI strategy” is making many capable leaders feel further behind than they actually are.

Every week brings another announcement about autonomous agents, predictive systems, synthetic content, or a new platform that promises to transform the entire revenue engine. For a marketing or sales leader already managing pipeline targets, campaign deadlines, team development, reporting, and customer expectations, that language can make AI feel like one more enormous initiative competing for attention.

The good news is that meaningful AI adoption does not have to begin with a company-wide strategy. It can begin with one frustrating task. My suggested approach and mindset is to warm the bay before boiling the ocean.

Start with friction, not technology

The most productive first question is not, “Where should we use AI?” It is, “What recurring work consumes time without requiring our best judgment?” That shift matters. It keeps the conversation grounded in actual work rather than novelty. It also keeps the focus on how you and your team leverage AI as support for your experiences and knowledge. 

Look for tasks that are repetitive, text-heavy, pattern-based, or slowed by a blank page. Examples include summarizing a sales call, converting interview notes into campaign themes, drafting subject-line options, comparing landing-page messages, creating a first-pass account brief, extracting action items from a meeting, or repurposing a webinar into social content. 

None of these activities requires you to redesign the department. Each gives a capable person a stronger starting point. That will be the common theme: I am seeing the best returns from AI, at the individual level, when you take a well-versed, knowledgeable, and experienced person and exponentially amplify their output.

Use AI for the first draft, not the final decision

The most attainable model is simple: You “train” or build the knowledge base for the AI, AI produces a first pass; a person improves and approves it.

That distinction removes much of the fear surrounding adoption. The marketer is still responsible for the insight, positioning, brand voice, factual accuracy, and final choice. The salesperson is still responsible for understanding the account, reading the room, building trust, and deciding what to say. AI reduces the time spent assembling raw material so the person can spend more time applying judgment.

This is leverage, not replacement.

A simple four-step starting framework

  1. Choose one recurring task. Pick something the team performs every week and understands well enough to evaluate. Make sure it is discreet with a start, end, and known expected outcome. 
  2. Define the input. Decide what information AI will receive and what information must remain outside an unapproved tool.
  3. Define the expected output. A vague request produces a vague result. Specify the audience, objective, format, tone, and constraints.
  4. Keep a human review step. Someone must verify the output, improve it, and remain accountable for anything used internally or externally.

Measure saved effort and improved outcomes

The first measure does not need to be sophisticated, but make sure you have one. It could be as simple as asking, did the task take 20 minutes instead of an hour? Did the team produce more useful options? Did the salesperson enter the call better prepared? Did content move from interview to publication more quickly?

As the workflow matures, connect it to a business outcome: response rate, meeting conversion, content production velocity, cost per opportunity, or time returned to the team. AI becomes valuable when it changes the work or the result—not when it merely creates more output.

The first AI plan can fit on one page

You do not need a steering committee, a new department, or a twelve-month roadmap to learn where AI can help. You need one useful problem, one approved tool, a clearly defined workflow, human review, and a measurable outcome.

Solve one frustrating task. Document what worked. Improve the instructions. Share the learning. Then choose the next task.

That is not a temporary experiment. It is how practical AI capability is built, one attainable workflow at a time. With that being said, make sure you are educated and careful with regards to data integrity and security. Make sure you research public LLMs vs. private or subscriptions. Depending on the information that will be shared with the technologies, you must be responsible.

Need help identifying the first three AI workflows that could create real leverage for your marketing or sales team? TruNorth Advisors can help you map the opportunities, guardrails, and measures that make adoption useful rather than overwhelming.