An AI checklist for a marketing team has 15 items, and the order matters more than any single item. Most teams start with tools. The data says that’s backwards. BCG found that 70% of AI success comes from people and processes. Only 10% comes from the technology itself. So that’s how this checklist is ordered: people first, tools later, measurement at the end.
The honest starting point: your team is already using AI
MIT’s NANDA lab found that 90% of workers already use personal AI tools every day. But only 40% of companies have official subscriptions. Your team is probably using ChatGPT, Grammarly, or Canva’s AI features right now. On their own accounts. Without anyone tracking it.
The industry calls this “shadow AI.” I call it the thing that should make you hurry.
Salesforce surveyed 4,450 marketers and found 75% had adopted AI. Sounds great until you read the next line: 84% of them still run generic campaigns. They have the tools. They don’t have the system. They can generate a campaign with AI in minutes, but without a checklist underneath, the output just sits there.
Jasper’s 2026 report found something even more surprising. The share of marketers who can prove AI ROI actually dropped from 49% to 41% year over year. More tools, less proof. So the real question isn’t “should we start using AI?” It’s “are we getting anything from the AI we’re already paying for?”
That’s what this checklist is for. Not starting from zero. Starting from the mess you already have and turning it into something that actually works. (If you haven’t identified your priorities yet, an AI assessment helps you figure out where to point AI before you start checking boxes.) If you’re still wondering whether AI marketing is legit, the data says yes, but only if you set it up right. If you want the bigger picture on where generative AI for marketing is headed, that post covers it. This one gives you the to-do list.
My take: most “AI readiness” guides assume you haven’t started yet. That’s fiction. Your team started months ago. The checklist is about catching up.
The 15-point AI checklist for marketing teams
This follows BCG’s 10-20-70 rule: 70% of success is people and processes, 20% is technology infrastructure, and 10% is algorithms. Most companies spend 80% of their AI budget on the 30% that matters least. The checklist flips that.
People and skills (start here)
1. Take inventory of what your team already uses. Send a five-minute survey. Ask everyone which AI tools they use for work, how often, and for what. You’ll be surprised. The point isn’t to police anyone. It’s to see what’s already working so you can build on it instead of starting over.
2. Pick one AI point person. This doesn’t need to be a data scientist. It’s the person on your team who’s most curious about AI and willing to stay current. Their job: test new tools, answer questions, and keep the team from all buying different subscriptions for the same thing. Jasper found that 65% of marketing teams now have a designated AI role. If you don’t, you’re falling behind.
3. Run a basic training session. Even one hour matters. Cover what AI can and can’t do. Show how to write a decent prompt (that’s just the instruction you give the AI). Point out where your team’s workflows could benefit. For the six patterns that handle most daily work, grab the AI prompt cheat sheet. The goal isn’t expertise. It’s getting everyone to the same starting line. Prosci’s research shows 43% of AI adoption failures come from weak leadership support, and 38% from not enough training.
4. Define who reviews AI output. Every piece of AI-generated work needs a human check before it goes out. Every single one. Decide now: who reviews AI drafts? Who approves AI-generated images? Who checks data pulled from AI tools? This isn’t bureaucracy. It’s quality control. For examples of how teams actually use AI in marketing, that post shows the full range. And for examples of AI in marketing with real company case studies, that post covers what worked and why.
Data and tools
5. Audit your data sources. List every tool that holds customer data. Your CRM (the system that tracks customer relationships), your email platform, your analytics, your ad platforms. Write them down. Note which ones talk to each other and which ones sit in silos. You can’t use AI effectively on data that’s scattered across ten tools that don’t connect. Gartner found that only 28% of AI projects fully succeed, and bad data is the top reason.
6. Clean your data. Remove duplicates in your CRM. Fix formatting in your email lists. Connect the systems that should be talking. This is the boring step that everyone skips. Don’t skip it. The AI-ready data checklist gives you the full four-check process. Salesforce found that teams with unified data are 42% more likely to deploy AI effectively. Forty-two percent. That’s not a rounding error.
7. Pick one tool for one workflow. Not five tools for five things. One tool. One workflow. Start with whatever task eats the most time and requires the least judgment. Email subject lines, social media captions, meeting summaries, report formatting. If affiliate content is your main channel, the AI affiliate marketing workflow is a good model for what “one workflow, done right” looks like. If you’re rolling out generative AI, content and research tasks are the best starting point because you can check quality in hours. If video is your bottleneck, start with AI video for marketing — the tools have gotten surprisingly good. If paid ads eat your mornings, see how to manage PPC with AI — bid management is one of the clearest automation wins. When you’re ready to compare options, start with choosing the right AI platform. Then check the best AI tools for marketing for the short list. If a website is on your checklist too, see real AI website builder examples to know what these tools actually produce.
8. Set up a shared prompt library. A prompt library is just a shared doc with the instructions that work well. When someone on your team figures out a prompt that writes great email subject lines, save it where everyone can find it. Stop every person from reinventing the wheel every morning. An AI assistant for your business works much better when everyone feeds it consistent instructions.
Policies and guardrails
9. Write a one-page AI use policy. Keep it simple. One page, three sections: approved tools, data rules, and what needs human review. That’s it. If your policy is longer than one page, nobody will read it. The NIST AI Risk Management Framework is the gold standard for larger companies, but for a small marketing team, one page is enough.
10. Define data boundaries. Some data should never go into an AI tool. Customer names, financial data, proprietary strategy docs, anything covered by privacy regulations. Write down the list. Share it with the team. Make the line clear so nobody has to guess.
11. Set brand and voice guidelines for AI content. AI doesn’t know your brand voice unless you tell it. Create a short reference doc: how you sound, words you use, words you avoid, example sentences. Paste it into every prompt. This one step is the difference between AI content that sounds like you and AI content that sounds like everyone else. It’s also worth managing your AI reputation, because chatbots and AI search engines are already describing your brand to potential customers whether you’ve guided them or not. For teams using AI heavily in content, AI-enhanced content marketing goes deeper.
12. Create a review workflow. Who checks what, and when? Map it out. Maybe the AI point person reviews AI tool output weekly. Maybe your editor approves all AI-generated blog content before publish. The format doesn’t matter. Having one does.
Measurement
13. Define what “working” looks like. Pick one metric for your first AI use case. Time saved per task. Output volume per week. Quality score from your review process. Just one. If you can’t say what success looks like, you can’t tell whether you got it. Jasper found only 41% of marketers can prove AI ROI. Don’t join the 59% who can’t.
14. Track both speed and quality. AI that’s fast but produces garbage isn’t saving you anything. Measure both. If your team writes 3x more blog posts but the quality drops, that’s not a win. It’s more work for whoever has to fix them.
15. Schedule a 30-day review. Put it on the calendar right now. In 30 days, sit down with the data and decide: expand this to another workflow, adjust how you’re using it, or stop. This review is non-negotiable. It’s what separates teams that improve from teams that just keep paying for tools they don’t use. Once you’ve done this review, an AI audit checklist helps you go deeper on what’s working and what isn’t.
What order to tackle this
The reason for this order is simple. PwC’s 2026 study found that technology delivers only 20% of AI’s value. The other 80% comes from redesigning how work actually gets done. So if you spend your first month shopping for tools, you’re working on the smallest part of the problem.
Week one: items 1 through 4. Survey the team, pick a point person, run a training session, decide who reviews output. This costs nothing but time.
Weeks two and three: items 5 through 8. Audit your data, clean it, pick one tool, build the prompt library. This is where you might spend some money, but keep it small.
Week three: items 9 through 12. Write your policy, define boundaries, set voice guidelines, create the review workflow. These take a few hours, not a few weeks.
Week four: items 13 through 15. Define success, start tracking, put the 30-day review on the calendar.
Four weeks. Not four months. The barriers to AI adoption that stop most teams aren’t technical. They’re organizational. This order tackles the organizational stuff first. Want a phased roadmap (crawl, walk, run)? That’s what the AI adoption framework is for. This checklist is the tactical companion: what to do this month.
My take: I’ve seen teams spend three months evaluating tools before anyone on the team could write a decent prompt. That’s like buying a gym membership before you own sneakers.
Three mistakes that stall most marketing teams
Trying everything at once. The RAND Corporation found that 80% of AI projects fail to deliver their intended value. For small teams, the reason is usually scope creep (starting with one thing, then adding five more before the first one works). Pick one workflow. Get it working. Then add the next. That’s not timid. That’s how the 6% who get real value from AI actually got there.
Skipping the data step. Between 48% and 52% of organizations name data quality as their number one AI adoption barrier. McKinsey’s data is clear: 88% of companies use AI somewhere, but only 6% are high performers. The gap is almost always dirty data or disconnected systems. Cleaning data is boring. It’s also the single highest-return activity on this entire checklist.
No measurement plan. If you can’t measure whether AI is helping, you can’t defend the budget when someone asks. And someone will ask. The Jasper stat is worth repeating: only 41% of marketers can prove their AI ROI. The other 59% are one budget review away from losing their tools.
If you want to implement AI in a single workflow properly, that post walks through the full process. This section is about the traps that catch teams who skip ahead.
Walk through it with someone who’s done it
Every marketing team is starting from a different place. Some have great data but no AI skills. Some have a team full of ChatGPT enthusiasts but zero governance. The checklist is the same for everyone. The priority order within it depends on where you are.
If you want to walk through where your team stands and figure out which items to tackle first, that’s exactly what the 15-minute spar is for. No pitch, no deck. Just an honest look at where you are on the list and what to do next.
FAQ
What is the 30% rule for AI?
The 30% rule is a rough guideline for splitting work between AI and humans. AI handles about 70% of the repetitive, time-consuming parts (drafts, data processing, scheduling). Humans keep the remaining 30% for judgment, creativity, and final review. It’s not a law. It’s a useful mental model for deciding what to automate and what to keep.
What is the 10-20-70 rule for AI?
BCG’s framework for where AI value actually comes from: 10% is algorithms (the AI models themselves), 20% is technology infrastructure (platforms and data), and 70% is people and processes (training, workflows, change management). The takeaway: most companies overspend on technology and underspend on the people side. If your AI budget is 80% tools and 20% training, flip it.
What is the best AI checklist app?
For making checklists with AI (like to-do lists or project plans), tools like Checklist.gg and Notion AI work well. But for the AI adoption checklist itself, you don’t need an app. You need the list above and one person accountable for each item. A shared spreadsheet with 15 rows, an owner column, and a done/not-done column is more than enough.
How long does it take to work through an AI checklist?
For a small marketing team (2 to 10 people), four to six weeks is realistic. People and skills items (1 through 4) can happen in week one. Data and tools (5 through 8) take two to three weeks. Policies (9 through 12) and measurement (13 through 15) fill the final weeks. The biggest variable is data cleanup. If your data is already clean, you’ll move faster.
Do I need a data scientist to use AI in marketing?
No. Modern AI marketing tools (ChatGPT, Claude, Jasper, HubSpot AI) are built for marketers, not engineers. What you need is clean data, clear policies, and someone willing to learn. The best AI tools for business are designed to work without technical expertise. The bottleneck is almost never technical skill. It’s willingness to change how you work.