AI Adoption Guidlines
AI Tools and features are exciting; you need some guidelines to be successful.
In this Executive Brief:
As this arrives in your inbox, we’re just a couple days from a TON of AI announcements at HubSpot’s annual InBound conference. There will be a lot of buzz and excitement, but you need to keep your head about you. To be successful you can’t just “Do AI”. Here are FranDev Lab’s Guidelines for successful AI projects.
What you need to know
There is a lot of excitement (and pressure) to adopt new AI tools and features. Franchisors should absolutely be looking for ways to improve their businesses with AI. Turning on every new AI feature and hoping it will solve all your problems is not a strategy for success.
Dedicated, thoughtful projects to use AI to improve what works in your business is a strategy for success.
You should get excited, but you should also be smart.
One thing you should take away from this is that AI is a tool. Like any tool, you need to be in control and know how to use it to get good results.
The Guidelines
ABC - Always Bring Context
In AI, Context is what we call the things you bring to the table to make AI work for you. AI will assume anything you don't tell it and AI doesn’t “understand” your business. The more blanks AI has to fill in, the worse its outputs will be. If 20 people in your business are using AI differently, with different context or no context, the outputs and outcomes will be further from the truth of your business.
You need to define context at each level of your business and for each team. Your team members need to know this and be equipped with your business’ shared context and their unique context to be successful.
This includes:
Your franchise’s products and services
Your franchise’s Target Market
Your franchise’s Brand Tone and Voice
The unique Policy and Regulatory constraints of franchising
Each teams’s goals, focus areas, processes, and methodologies
Each team member’s Role, Responsibilities, priorities, and constraints
Each task’s outcomes, requirements, and inputs
Keep your tools where you work
There’s a reason you don’t keep your hammer in the refrigerator. We keep our tools where we use them. AI tools that work where your team works, with consistent context, and with good training, are more secure, save more time, and produce better outcomes.
This means:
You should prefer native tools in the systems you use when have the choice.
It’s better if your team doesn’t have to stop and switch tools to use AI.
External tools always risk your data ending up somewhere it shouldn’t.
Native tools are preferred over integrated tools and both are preferred over going out to an external chat tool.
Sculpt your data and tell AI what matters
AI doesn’t really “understand” what’s important in your data or your business unless you tell it. You need give AI data that is sculpted to focus on what’s most important. You can feed AI a lot of data, but it won’t be able to figure out what is most important. It can tell you what’s there but not what isn’t there that should be.
If you don’t sculpt your data, every single AI request will be no different than asking a brand new employee with no information or training to complete a task. You wouldn’t ask that new employee to wade through a mess of data and figure out for themselves what matters and you shouldn’t do that with AI either.
This means:
You can’t just hope AI will “figure out” what matters.
You need to collect the data that’s important to your franchise in the systems where you’ll use AI.
That data needs to be reliable.
Your team needs know when and how to bring that data to the table.
Doing this well informs the
contextyour teams will use.
So, now what?
How you put these guidelines into practice will depend on your tools, your franchise, and how prepared your team is to work with AI. Any time you start an AI project, big or small, you should be thinking about these things.
And of course FranDev Lab is always here to help. I’m happy to talk about this any time, even if you just need some perspective or someone to talk it out with.
See you in the next Lab Report.