Key AI lessons for real world SMBs and mid-market marketers

In my recent Human First AI Marketing conversation with AI futurist and author David Espindola, we dug into what the development of AI means for small and mid-market companies that need to win customers this quarter, and not just in some distant future. We talked about why human and machine intelligence complement one another, the importance of differentiating an AI voice from real relationship, and what a good AI roadmap looks like for businesses that do not have a billion dollar R&D budget.

Here is the short version. Let bots handle the repetitive, digital, rule based work; you and your team handle the creative, relational, and trust building work. That simple split runs through everything David shared, from his makers–shapers–takers framework, to his work with the U.S. Small Business Administration, to his vision of a future where AI handles the chores and we humans get more time to focus on what inspires and fulfills us.

1. Human first really means human across the whole workflow

David and I both like the phrase “human in the loop,” although in practice it means human at the beginning, human in the middle, and human at the end.

• At the beginning, humans set direction; you define the problem worth solving, pick the use case, and decide what “good” looks like.
• In the middle, humans monitor and steer; you check whether AI is behaving, staying on brand, and staying inside your ethical lines.
• At the end, humans own the output; you decide what gets published, shipped, or sent to a customer.

AI can sound extremely confident while being spectacularly wrong. That combination looks impressive on a demo reel and dangerous in front of your customers. You need people with judgment in every loop, especially in marketing and sales.

2. Context training beats copy paste prompts

One of David’s strongest points; generic AI gives generic answers.

Employees dropping product requests into a free chat tool and copy pasting emails to customers will eventually burn the brand. The model does not know your pricing, your positioning, or your quirky sense of humor. It only knows “professional email voice.”

The shift that matters for SMBs and midmarket teams:

• Feed AI your content, FAQs, product docs, and brand guidelines.
• Give it examples of “this sounds like us” and “this does not.”
• Use retrieval or fine tuning so your tools answer based on your material first.

That extra context sharply reduces hallucinations and makes the AI sound like part of your company instead of a robot intern who wandered in from the internet.

3. Pick your lane: maker, shaper, or taker

David shared a simple strategy lens that I really like for business leaders.

• Makers build AI itself; huge R&D budgets; high risk and high reward. If your company is not already in that world, you can safely skip this lane.
• Takers adopt AI that is baked into their existing tools; think Microsoft Copilot, Google Gemini in Workspace, AI add ons in CRM and marketing platforms. You stay current; you do not try to win with AI as your main differentiator.
• Shapers live in the middle; they use existing models, then shape them with prompts, custom data, agents, and integrations to fit their business.

Most SMBs and midmarket companies will do well as smart shapers with a bit of taker behavior. You can let your vendors carry the heavy engineering load, then shape AI around your workflows and your customers. That beats chasing the hype cycle or pretending AI will blow over.

4. Start with problems, not with tools

David’s number one recommendation; list real business problems first, then match AI to the ones where it naturally shines.

Good candidates for AI help in SMB and midmarket marketing and sales:

• Responding to repeat questions in support and sales.
• Summarizing calls, emails, and chat transcripts into CRM notes.
• Drafting first versions of blogs, emails, and ads in your voice.
• Generating variations for A/B testing and personalization.
• Spotting patterns in customer behavior or churn.

Weak candidates: anything that lives mostly in relationship, trust, and high stakes judgment. AI can help with research and prep there; it should not close the deal or approve the contract.

5. Hyper personalization without losing your soul

David predicts a serious rise in hyper personalization; offers and experiences tuned to very small segments or even individuals. AI finally gives smaller teams enough horsepower to do that kind of work without hiring an army.

The risk; you can accidentally optimize your way into bland, repetitive content that “works” in the dashboard yet bores real humans. The opportunity; use AI to handle the complexity and heavy lifting, then spend your human time on big original ideas that break patterns and create emotional connection.

Think of AI as your analytics and variation engine. Humans still own the story and the heart.

6. Sales teams finally escape CRM jail

David has lived in the enterprise systems world for a long time and made a great point; CRM made managers happy and salespeople tired.

AI can flip that.

• Voice or meeting AI can log activities and summarize calls automatically.
• Smart assistants can extract next steps and update deal stages.
• Salespeople can walk out of a meeting, talk through what happened, and let AI convert that into clean CRM data.

Nobody got into sales because they love dropdown menus. When AI does the logging and formatting, reps can spend more time face to face or screen to screen with actual customers. That is where trust lives and that is where big deals close.

7. Guardrails, ethics, and the Zena effect

David hosts a show with an AI co host named Zena. It is fun; it is also a live example of why we need guardrails. These systems mimic human voices and emotions with scary accuracy. They do not have consciousness, feelings, or a “soul,” no matter how charming the agent voice sounds.

For your company, that means clear boundaries:

• Always disclose when someone is interacting with an AI agent.
• Decide what decisions AI can never make on its own.
• Set rules for data, privacy, and model access.
• Keep a human escalation path easy to reach.

As AI becomes part of more customer experiences, trust and transparency become real assets, not just nice slogans.

8. What this means for your team and your job

Short term, AI will automate specific tasks; long term, anything that is digital and rule based will keep moving in that direction. David expects trades and complex physical work to remain human heavy for a long time. He also expects human skills like creativity, empathy, and ethical judgment to increase in value as machines handle more of the repeatable stuff.

For marketers and salespeople, that means:

• Less time formatting, drafting, and logging.
• More time listening, thinking, and connecting.
• Stronger demand for people who can steer AI and tell powerful stories.

AI does not remove the need for human work; it changes the kind of work that matters.

9. A simple Human First AI action plan for SMBs

If you run or market an SMB or midmarket company, here is a practical starting checklist based on David’s advice and our conversation:

1. Decide whether you are mostly a maker, shaper, or taker.
2. Nominate an internal AI champion who is already a bit ahead of the curve.
3. Run a short workshop to list your top 10 problems or bottlenecks.
4. Pick 2 or 3 AI friendly use cases and design human in the loop workflows.
5. Train your tools with your data, your content, and your brand voice.
6. Write down a simple AI policy that covers ethics, approvals, and data use.
7. Measure results, keep what works, then expand from there.

You do not need to rebuild your whole company in one sprint. You only need to keep moving, stay human at the center, and let the bots take the busywork off your plate.

If you want help figuring out what that looks like for your marketing and sales team, Avenue9 lives in this space every day.