1838% ROI on AI Tools (Part 2 of 2)

A little note: Firstly…this newsletter is 6 days late, many apologies. Monday brought with it a pretty horrible medical emergency with my husband. I managed to get the email version out on Wednesday, but then it all kicked off again. It's been a rough week, but he's back home now and it's time to get caught up.

Here we are at Part 2 of my two-parter, “1838% ROI on AI Tools”. I’ve had great feedback on the first part, so I’m hoping you’ll find the second half just as fruitful.

Last week, I took you through the first 5 steps:

Step 1: Look at business/role goals

Step 2: Explore the obstacles in the way

Step 3: Use the Impact/Effort Matrix to prioritise them

Step 4: Prompt Perplexity to see which can be solved by AI

Step 5: Search for the right AI tools

This week, we’ll do the final 4 steps:

Step 6: How to assess and pilot AI tools

Step 7: Critical questions to ask around data privacy

Step 8: The whole AI integration framework broken down visually

Step 9: My own ROI breakdown (1838%)

Before we get into that, the very quickest of notes :

⏰ 1 day left to buy tickets!

On Tuesday, my very first Build an AI-First Business bootcamp starts, so if you’re looking to join us, this is your last chance to grab yours.

6 quick bootcamp facts:

• It runs over 4 days, 22-25 October
• Each session is 2 hours 4:30-6:30pm BST / 11:30 - 1:30 ET
• It’ll be theory, interactive activities and homework tasks
• The bootcamp is aimed at ambitious leaders wanting to integrate AI
• Recordings available to all ticket holders (released the following week)
• It’s only £72 / $97 per ticket, including lifetime access to 2 exclusive groups

Can’t make it? Don’t worry, here is my upcoming bootcamp and masterclass schedule (bootcamps run over a few days, masterclasses are one intensive blast):

• Supercharge Your Sales with AI (November 2024)
• Build Your 2025 Marketing Strategy with AI (December 2024)
• Build an AI LinkedIn Post Machine (December 2024)
• Create Killer AI Content (Text/Image/Video/Audio) (January 2025)
• Become an AI Trainer (February 2025)

Join the waiting list here to be told when tickets get released for each bootcamp, before anyone else.

Right, let’s get on with it, shall we?

There are hundreds of thousands of AI tools out there - well, I’ve no idea exactly how many, but it certainly feels that way. Even as someone who works full time in the generative AI world, I’m still completely overwhelmed for choice. Every day, there’s something new.

Last time, I introduced you to how to find AI tools using sites like FutureTools.io, but let’s say you’ve found a few interesting tools, and you want to check them out. What next?

Step 6: How to Properly Assess Tools

You’ve got your list of potential tools - great. But before you go all in, you need to put them through their paces. For this, you’ll need my IMPACTT framework:

IMPACTT Framework

I - Integration

- How well does the tool integrate with your existing systems and workflows? Does it fit nicely alongside them? Does it replace several older tools? Does it have APIs (these help you link tools and softwares together)?

- How easy is it to get set up? (Don’t forget to include any data you need to feed into it).

M - Metrics

- What key performance indicators (KPIs) will you use to measure the tool's success?

This could include time saved, accuracy improvements, increased depth of service, or cost reductions. Many of my clients also measure team enthusiasm and uptake.

P - Performance

- How well does the tool actually perform its intended function?

Test it against your specific use cases and compare it to your current human methods. Look at accuracy and consistency, as well as speed and effectiveness.

A - Adaptability

- How flexible is the tool?

- Can it handle a variety of inputs and scenarios?

Consider how well it might adapt to future needs and how fast AI is evolving. Are they regularly releasing updates and new versions to stay ahead?

C - Cost

- What's the total cost of ownership?

First calculate how many hours this could save you (see my ROI video later in this newsletter for my calculations) (A).

Then add up tool subscription fees, implementation costs and training expenses (B).

If A is more than B, you may have a winner.

T - Training

- How much training is required for your team to use the tool effectively? - Do they have a basic understanding of generative AI yet?

Consider the learning curve properly and evaluate available support resources.

T - Trust & Security

How reliable and secure is the tool? Dig into their policies on the site - absence of these is a HUGE red flag.

See below for questions to ask about security, direct from a tech lawyer specialising in AI.

IMPACTT Framework (AI for Non-Techies 2024)

Let’s Dig More Into Security

When it comes to security, AI is the wild west right now. So I asked a tech lawyer and two AI tool founders about the sort of things we should be asking.

Here are the results:

Here are a few other questions I’d ask:

• Where is our data stored, and who has access to it?
• How is our data used to train the AI models?
• Can we opt out of having our data used for training?
• What's the process if we need to delete our data?
• How does the tool comply with regulations like GDPR or CCPA?

Happy and secure? Let’s move on and actually use the thing.

Step 7: Organising a Pilot Program

You’ve vetted your tools, it's time to test them in the wild. Here's how to set up a strong pilot program:

1. Choose Your Champions: Pick a small, diverse group of team members who are excited about AI. Their enthusiasm will be infectious. They don’t have to be techie types (I mean, look at me!), they just need to be interested.
2. Set Clear Goals: What does success look like for this pilot? Be specific and measurable. Regularly check in against these goals.
3. Provide Training: Don't just throw people in the deep end. Give them the knowledge they need to succeed- work with top notch AI trainers.
4. Start Small: Always begin with those smaller, quick-win tasks we drafted in Step 3. As your confidence grows, you can tackle those bigger, strategic tasks.
5. Gather Feedback: Regular check-ins are crucial. What's working? What isn't? What unexpected uses have people found?
6. Iterate and Improve: Use the feedback to refine your approach. AI integration is a journey, not a destination.

Step 8: The AI Integration Framework

Alright, steps are helpful, but it’s easier to see in one place.

Let's pull all of this together into a visual framework you can revisit and use:

Note: People always ask if they can use my models and frameworks in their own trainings - that’s absolutely fine. Please just credit AI for Non-Techies and link to my site please.

And Finally, My 1838% ROI Breakdown

Now, I promised you a breakdown of my own 1838% ROI, so here it is.

If you follow me on LinkedIn, you may already have seen my post sharing this a week or two ago. But it’s worth seeing again - as I think it’s a simple but clear way to show ROI on AI, and I haven’t seen anything like it before or since.

I wrote down all of the tools and use, and estimated the hours they saved me each month. (An educated guess, from someone who’s done all these things manually before for years).

I’ve got a real mix of hourlys - my consultancy hourly is dramatically different to my training hourly, and that’s different to my everyday admin hourly. So rather than take my wonky hourly rates, I took the UK average hourly rate of £17.48, for the sake of balance.

So that’s 194 hours saved monthly x £17.48 = £3,391.12

Less cost of tools = £175/month

= ROI of 1837.7% (Nope, that’s not a typo.)

PLUS - that doesn’t factor in:

→ The new revenue generated by these tools → Tools I use to take care of my mental health → New services I can offer via these tools

DOUBLE PLUS - did you notice that 194 hours is basically a whole full time employee? It's a very exciting time for small businesses like mine, who can get to higher revenue levels than ever before without hiring.

And all that is just AI alone - imagine the ROI with automation, too. Cor.

Wrapping Up

There you have it, folks – your two-part roadmap to AI integration success. I’d love to know what you think, and learn about the results you get from implementing it.

Remember, the key is to start with problems, not technology. Stay focused on your goals, be methodical in your approach, and always, always prioritise data privacy and security.

Next week (Monday 28th, I need a week off to catch up after a week of horrible hospitals), we’ll be exploring the art of cloning yourself with AI using 3 cutting-edge tools: Veed, ElevenLabs and Coachvox.

Have a wonderful week,

Heather