We know we should use AI, but it isn't paying off

More subscriptions don't automatically mean less work, whether they are AI tools or any other software. Tools pay off when they're tied to a specific bottleneck and you can show the result.

Try this first

1. List every software subscription you pay for, AI or not, starting with your three most expensive.

2. Write down what each one was supposed to save: time, money or headcount.

3. Check whether it delivered that, in a way you could show someone.

4. Cancel or rethink anything that fails. Next time, decide how you'll measure success before you choose the tool.

How you'll know it's working: you can point to a specific task that now takes less time, and you're paying for fewer tools.

Go deeper

• You Bought the AI Tools. So Why Is Everything Still Taking Longer?

• Why Most AI Marketing Advice Is Setting You Up to Fail

• AI Without Strategy Is Just Expensive Automation

• The Lazy Engineer's Advantage, Chapter 8

Whether AI pays off depends more on your readiness than on the tool you choose. Two things matter most: whether your information is usable, and whether your team has the time and skills.

Try this first

1. Rate your data: is the information AI would need organized and easy to reach?

2. Rate your team: does someone have the time and skills to learn and use it?

3. Use those ratings to choose your next move. Strong on both: go ahead. Good data, thin team: start small. Capable team, messy data: fix the foundation first. Weak on both: AI isn't your next step yet.

4. Tie your first project to one marketing or business goal you already have.

How you'll know it's working: your first project has a clear goal, an owner and a way to tell whether it worked.

Go deeper

• Most AI Marketing Advice Assumes You're Ready. Here's How to Find Out If You Actually Are.

• Why Your Marketing Strategy and AI Strategy Must Be One Strategy

• The AI Strategy Conversation Your Leadership Team Keeps Avoiding

• The B2B Marketing Reality Check, Chapters 12 and 13

You can start with one task today. The free versions of the major AI tools are enough for these, and you'll stay in charge of the final decisions.

Try this first

1. Before your next meeting, have AI summarize the background documents into key points and questions.

2. Paste in customer comments or survey answers and ask for the main themes.

3. Have it write a first draft of something routine, then edit it instead of starting from a blank page.

4. Track the minutes saved, check the results before you use them, and keep confidential information out of tools your company hasn't approved.

How you'll know it's working: a task you do every week takes noticeably less time, and the quality holds up.

Go deeper

• 3 AI Quick Wins You Can Implement Before Your Next Team Meeting

• 4 AI Quick Wins That Turn Customer Data Into Marketing Gold

• 3 AI Quick Wins That Give Your Marketing Team 10+ Hours Back Per Week

Automation money tends to follow the loudest complaint, not the most costly process. Score the work first, and treat automation as the last option after simpler fixes.

Try this first

1. List your recurring tasks and score each from 1 to 5 on three things: how error-prone it is, how much manual effort it takes and how often it happens.

2. See whether you can eliminate, standardize or template the work before buying anything.

3. Look at the pattern. High effort and frequency with few errors is the best candidate. Lots of errors means standardize first.

4. Keep a person on the task if you can't write down the rules, the math doesn't work or the volume is too low.

How you'll know it's working: 90 days after a change, you can show the time or errors it saved.

Go deeper

• Before You Automate Anything, Run This 3-Signal Triage (includes a free calculator)

• The Lazy Engineer's Advantage, Chapters 2, 3, 9 and 11 (the triage, the decision framework, when not to automate, and auditing your week)

Do we need paid tools to start?

No. Free versions are enough for your first experiments. Pay only when you've found a use that justifies it.

Will AI replace our marketing?

It can speed up the work, but deciding who you serve and what to say still takes human judgment.

We've already spent money on tools that didn't help. Now what?

Review what you have, cancel what isn't delivering, and start again with one specific task.