May Contain Jab
← All episodes

Episode 11

How to Pick the Right AI Tools for Your Business (Without Wasting Money)

October 27, 2026 ·30 min

Before you pay for another AI licence, run this audit. Jab and May share the six-step process they use to pick AI tools for small businesses

Show Notes

How to Pick AI Tools

Two separate pieces. The show notes go on the episode page (the Sanity episode document) under the article in the Sanity pack. The blog post is a standalone post with its own slug that links back to the episode, written from a different angle so the two pages don't compete in search. UK English, no em dashes, every claim from the episode. Quote timestamps are from the raw recording.

Part 1: Show notes

Episode summary

There's an AI tool for everything, so which one is right for your business? Jab and May take a $5 million business running on legacy marketing and sales systems, with no SOPs and no AI beyond basic ChatGPT, and walk through the full audit they would run before recommending a single tool. From team beliefs and tool overlap to Porter's five forces, the SAF test and the Deming cycle, this is the thinking to do before you pay for another licence.

What you'll learn

  • How to audit your business inside out: people, tools, data, SOPs and budget
  • How to decide whether to keep, cull or retrain on the tools you already pay for
  • Why leadership buy-in and change management make or break adoption
  • How suppliers, substitutes and regulation should shape your choice
  • How to test options for suitability, acceptability and feasibility
  • Which criteria to measure a tool against, from ROI to customer effort score
  • Why implementation is only the start, and how the Deming cycle keeps you improving

Timestamps

  • 00:00 Welcome: solving real problems
  • 02:22 Audit first, like a doctor
  • 05:18 Tool audit: keep, cull, retrain
  • 07:26 Budget, gaps and skills
  • 09:26 Buy-in, change and suppliers
  • 11:32 Competitors and five forces
  • 13:50 Regulation and tech changes
  • 16:03 Your edge and SAF test
  • 19:06 Setting your decision criteria
  • 23:05 Three options, one decision
  • 27:06 Implementation and onboarding
  • 29:36 Deming cycle and takeaways

Quotes from the episode

  • "We're not gonna AI for AI's sake." Jab (02:36)
  • "If you're paying for these tools and you're not using them, there's no need to have them in your system." May (05:18)
  • "It's a different winner every day." May (15:16)
  • "There's a tool out there for everyone, for everything." May (15:20)
  • "If people are willing to use a spreadsheet, let them use a spreadsheet." Jab (25:44)
  • "Operational efficiency is not a one-time thing." Jab (29:46)

Frameworks mentioned

  • Porter's five forces: threat of new entrants, bargaining power of suppliers and buyers, substitutes, and rivalry
  • SAF test: suitability, acceptability, feasibility
  • Gap analysis: what capability is missing, not just what to cut
  • Deming cycle: plan, do, check, act
  • Metrics: ROI, customer lifetime value (LTV), customer satisfaction (CSAT), customer effort score (CES)

Tools mentioned

ChatGPT, Claude, Claude Code, Sonnet, Perplexity (May's pick for research), Manus, Adobe Premiere Pro and Sage. Nothing in this episode is sponsored.

Your turn

Pick the most expensive tool you pay for and ask three questions. Is everyone actually using it? Does it overlap with something else you have? Is it pulling its weight? Keep, cull or retrain: tell us which.

Connect

  • Email: mayandjab@maycontainjab.com
  • May Zhan (AI automation and strategy): https://linkedin.com/in/may-zhan
  • Jab Siwela (audience, resonance and revenue): https://linkedin.com/in/jab-siwela
  • Listen on YouTube, Spotify and Apple Podcasts (add links once live)
How to Pick the Right AI Tools for Your Business (Without Wasting Money)

Most owners have done it. You see a tool everywhere, someone you respect swears by it, and you sign up just in case. Then reality hits. Setting it up takes time you don't have, nobody on the team knows how to use it properly, and three months later you're paying for something nobody opens.

The tool usually isn't the problem. The problem is that nobody asked the right questions before buying it.

On a recent episode of May Contain Jab, we walked through how we'd choose AI tools for a $5 million business running on legacy systems, with no documented processes and no AI beyond basic ChatGPT. These are the 12 questions that came out of it. Answer them honestly and you'll know which tool you need, or whether you need a new one at all.

Your people

1. What does your team believe about change? Before you look at any software, look at the people who'll use it. Are they early adopters or laggards? A brilliant tool handed to a team that doesn't believe in it will gather dust.

2. Is your leadership bought in, or just you? If the founder is excited but the rest of the leadership team isn't, you have a problem. Leaders are the ones meant to rally everyone else. If they're not on board, you need a change management plan before you need a new tool.

3. Does your team have the skills, and will you upskill them? Every tool needs someone who can set it up and use it well. If the skills aren't there, budget for training as part of the decision, not as an afterthought.

Your processes and data

4. Are your processes written down, and actually followed? You can grow a business to a couple of million without writing anything down. You can't automate it that way. AI needs a process to follow. If yours lives in someone's head, document it first.

5. Is your data clean, and in one place? Messy, scattered data is one of the most common problems we find in businesses that haven't modernised yet. An AI tool working from bad data just produces bad answers faster.

Your existing tools

6. Is anyone actually using the tools you already pay for? Start with your most expensive tool and work down. Is it used to full capacity? Could you downsize the plan, or cancel it? For anything underused, decide whether to keep it, cull it or retrain your team on it.

7. Does the new tool overlap with something you already have? Overlap is easy to miss, and it's getting worse now that AI features are built into almost everything. Check whether something in your current stack already does the job before you add another subscription.

8. Will it integrate with your systems, and your suppliers'? Legacy systems that don't talk to each other are a hidden tax on every task. Look one layer out as well: will a new tool improve how you work with suppliers, or make it harder?

Your market

9. What are your competitors and substitutes doing faster? Go wider than your direct competitors. An energy drink isn't just competing with other energy drinks. It's competing with coffee and bubble tea too. If a substitute is quicker or easier, your customers will hold you to that standard. Porter's five forces is an old model, but it's still a useful way to see where you're being outpaced.

10. Does it keep you within the rules? AI will happily produce marketing that breaks regulations or platform rules if nobody checks it. If Meta or TikTok blocks your ad account, you can't reach your customers. Factor regulation into every tool decision.

The solution itself

11. Is it suitable, acceptable and feasible? Run every option through the SAF test. Is it suitable for the people who'll use it? The most powerful software isn't the right choice for someone who only needs the basics. Is it acceptable to senior management, staff and finance? Is it feasible for three to five years, not just 45 days? Think January gym resolutions that are over by Valentine's Day.

12. How will you measure it, and who's watching? Decide your criteria before you compare options: ease of use, cost, efficiency, ROI and how long results take to show. Some, like LLM SEO, can take three months to a year. On the customer side, ask whether it will improve lifetime value, satisfaction (CSAT) and customer effort score (CES). Then build in monitoring from day one. Tools don't run themselves, and if nobody is checking, you won't know when something breaks.

What to do with your answers

Once you've worked through the questions, shortlist the options that fit your budget, your needs and your people. Narrow them to three, pick one recommendation, and make the decision.

Don't be surprised if the answer is simpler than you expected. We've seen an AI-layered project management tool replaced by a plain spreadsheet, because the team would actually use the spreadsheet. If people will use it, that's often the right answer.

After you implement, keep reviewing. Plan, do, check, act, then go round again. Choosing tools isn't a one-off decision. It's a cycle.

Hear the full breakdown

We go through every step, with examples, in the episode How to Pick the Right AI Tools for Your Business. Listen on YouTube, Spotify or Apple Podcasts.

Feeling overwhelmed? Take a breath, then email mayandjab@maycontainjab.com and we'll help you think it through.

Never miss an episode.

Join the list for new episodes, behind-the-scenes notes, and early access.

No spam. Unsubscribe anytime.