Intro
So, you’ve been hearing about AI
everywhere. Your competitors are talking
about it. Your software vendors are
pushing it. And honestly, it feels like
if you’re not using AI, you’re falling
behind.
But here’s the thing. I work with small
businesses every week implementing
automation and AI solutions and I can
tell you right now most businesses jump
into AI without asking the right
questions first. They waste money. They
create more problems than they solve and
they end up more confused than when they
started. And perhaps this is why around
80% of AI projects are failing right
now.
So before you spend a penny on AI tools
or hire someone like me, you probably
need to answer these 10 questions first.
If you can’t answer at least seven of
them fairly clearly, you’re probably not
ready. But first, um, a quick reality
check. So you’ll have heard the phrase
crap in, crap out. If your processes
aren’t repeatable and documented right
now, AI is probably not going to fix
things for you. It’ll just amplify your
mess at scale. Uh, and I’ve seen this
happen. So, if your systems are chaos,
pause this video, get your workflows
documented and working consistently, and
then come back to it.
AI and automation are accelerators. So
they will generally speed up what’s
already working. They don’t create any
order from chaos.
Okay. So let’s go through the questions.
So question number one, what specific
problem am I trying to solve? So it
should not be I want AI to make things
better or I want AI I want AI my
business. This is just too vague.
Be specific. What exactly are you trying
to fix? Are you spending 10 hours a week
on manual data entry? Are customer
support emails piling up and taking days
to answer? Are you manually chasing
invoices?
Be specific. Write it down will help. I
want to reduce the time spent
categorizing support emails from 2 hours
a day to 50 minutes. This is more like a
problem AI can solve.
So if you can’t describe the problem
clearly, AI probably won’t fix it.
Question number two is this problem
actually worth solving with AI? So some
problems are just annoying, others cost
you thousands of dollars or hours of
time every month. Do the maths. If
fixing this problem saves you £500 a
year, but the AI tool costs £1,200 a
year plus 20 hours to set it up, then
it’s obviously not worth it. But if it
saves you 10 hours a week, now we’re
talking. So focus on your expensive
problems first. And that can be
financial or it can be time. Question
three,
do I already have tools that AI can
connect with? This is huge. Um, don’t
start from scratch if you don’t have to.
Check what you’re already using. Uh do
you have Pipe Drive, Asana, Gmail,
Notion, Air Table? Most AI tools can
plug into these platforms. Some have AI
already built in. So that means you’re
building on what you already know and
you’re not learning five new systems.
So when I set up AI for clients, I
always start with their existing stack.
It’s faster, cheaper, and way less
painful.
And a note of warning here is that not
every piece of software with an API will
plug and play. Sometimes there’s a lot
of potential software development to get
things working with that API. Question
number four, well, what’s the simplest
way to start?
You don’t need a developer. You don’t
need to build custom GPT agents. You
probably don’t even need me yet. Start
small. Maybe it’s using chat GPT to
draft your weekly emails. Maybe it’s a
simple Zapia automation that logs leads
into your CRM.
Maybe it’s an AI tool that summarizes
your meeting notes or uses existing
summaries of meeting notes. So, pick one
small annoying task and automate that.
See if that works and then build from
there. I always tell clients, start with
one workflow,
get it right, get it consistent, and
then scale. Don’t try and automate your
entire business on day one. Question
number five, how will I know if it’s
working? seems like an obvious question,
but if you can’t measure success, you
won’t know if AI is helping or getting
in the way. So, set a benchmark before
you start. So, right now, it takes me 5
hours a week to manually update a
project checker, and I want to reduce
that to 1 hour. Or our average email
response time is 24 hours, and I want to
get that to under four.
Track it. Measure it. If it’s not moving
the needle after a month, change your
approach or just ditch the tool. It’s
not working.
Question number six, who’s actually
going to manage this?
Um, here’s where a lot of businesses
trip up. They buy the tool, someone sets
it up once, and then 6 months later, no
one’s using it. Who on your team is
going to own this? Do they have the time
to learn it? Are they actually going to
use it? Or will it just sit there
gathering dust? If you don’t have
someone who can manage it, you’ll either
need to hire help or pick a simpler
solution. Training and adoption are not
really optional here.
Question number seven, how accurate does
this need to be? So for brainstorming or
for drafting social media posts, AI can
be sort of 80% accurate and that’s fine.
You will do the rest. But for invoicing,
payroll, legal contracts, you need 99.9
if not 100% accuracy. AI isn’t there yet
for some of the high stakes tasks
without some sort of human oversight. So
match the tool to the risk level or add
in a human approval. Don’t use AI for
anything where mistakes would cost you
clients or money or you would fail your
compliance. Question number eight, is my
data secure?
And this matters a lot more than most
people think. If you’re feeding customer
data, financial information, or anything
sensitive into an AI tool, check the
privacy policy and where the data is
stored.
Um, is it being used to train their
models? Are you violating your GDPR or
your own client contracts?
I’ve had to walk some clients back from
tours that look really great but had
terrible data security and it’s simply
not worth the risk. Question number
nine, how will this affect my team and
my customers?
Be honest. Will AI replace tasks? Will
it free up your team for more valuable
work? Or will it make people a bit
nervous about their jobs?
So, communicate clearly. Frame it as
this tool handles the boring stuff so
you can focus on what you’re actually
good at. So, bring your team into the
process. Let them test it, get their
feedback.
Same with customers. If you are using AI
for customer support, make sure it
actually improves their experience.
Don’t try and use AI to save money if it
makes your service worse. And you will
be surprised that some of the really big
companies are not covering this area.
Some of their service is now terrible
with AI.
Question number 10.
Can this scale with my business?
So what works for 100 customers might
not work for a thousand. What works for
five team members might break at 20.
Uh pick tools that can grow with you.
Check the pricing tiers. Make sure
you’re not locking yourself into
something that will be obsolete or
unaffordable in 2 years.
try and futurep proof your setup where
you can. So, here’s where you are. If
you can confidently answer yes to at
least seven of these questions, you’re
probably ready to explore AI for your
business. And if not, that’s okay. Start
small, pick one workflow, document it,
and experiment. If you’re stuck or you
want someone to help you figure this
out, this is literally what I do. I help
businesses implement AI and automation
into their workflows. So, if you want to
book a free call to talk to me, there’s
a link in the description below. So, if
this was helpful, let me know in the
comments below. Maybe even which of
these questions got you thinking. Uh,
and if you like this video, you may like
some of my other videos on AI and
automation, too. Thanks for watching,
and I’ll see you in the next one.