No guessing, no automate everything hype. Just a simple filter of five questions you can use.

And if the test passes all five, automate it. Use AI, go the whole nine yards. And if it doesn’t, leave it alone.

1. Do I do this more than once a week?

This is a big one because we want to automate the bottlenecks. We want to automate the things that are actually going to give you time back.

I would say around 60 to 65 maybe 70 percent of the workday is just repetitive boring tasks that you probably already know software can handle but you got to do it anyway.

And so what I would do is I would take the most repetitive tasks that you’re doing every day and try to document them. You want to take a look at this and then you want to list out all the tasks you do on a week-to-week basis. You want to rank them by which one you hate the most and which ones you hate the least.

Even if they’re spending an hour on it each day it’s going to be draining them and by the end of the day they’re going to be working less and being less productive than if you were able to tackle that same task automatically.

2. Are all the steps the same?

Where automation really thrives is when it’s given a list of tasks that it can do at each time.

Go ahead and take one of the tasks that you made from your first list. Try to write all the steps down on a piece of paper. If you get to a point where you’re saying it depends or it’s a completely different route, sometimes you can handle that with automation, but it’s not a priority, easy automation that you want to tackle first.

3. What happens when things go wrong?

What would the impact be to your business if this specific task went wrong?

You want this to be a low impact. You don’t want it to be costing your company a lot of money.

4. Can I measure the before and after?

There’s two verticals I look at. There’s the time saved vertical, which is really important. And it’s definitely a measurable thing. You just take what their time is worth per hour and you multiply it by however hours that task takes per week, per month.

So the other vertical is increasing leads, increasing income for your company. This might be a little trickier to put a number down to it.

So when I’m going in to measure the before and after, I always got to ask myself, is this A, saving X amount of hours and how much is X amount of hours worth? Or is it B, bringing in X amount of dollars because of the systems in place?

Let’s say you don’t have a lead problem at all. You just have a capacity problem. Let’s just say you’re able to handle 20 clients a month, right? If through automations and efficiency of your workflow, you’re able to get that to 40 clients a month, then you’ve effectively doubled your income stream.

If we can handle double the amount of clients, then we’re making double the amount of money without adding more people on payroll.

And it fails if you can’t define what better looks like.

5. Am I fixing a broken process or automating a working one?

This is a great one because it’s so true. If the process doesn’t even exist, then how are we going to even automate it?

It is not just yes or no

There’s three buckets that you want to think about.

There’s a fully automated bucket where you’re taking processes and they’re done basically the same every time.

You have AI assisted, which I’m calling workflow AI. And this AI fills in the gaps for when the steps aren’t the same each time and you need an AI to fill in the gaps, basically.

And of course, the third bucket will be human only. These are high stakes.

So it’s not just yes or no. There’s levels to this, and I think of it in three buckets.

The five questions

  1. Do I do this more than once a week?
  2. Does it follow the same steps every time?
  3. If it breaks and nobody notices, is the damage small or is it big?
  4. Can I measure what the cost is to me right now in dollars?
  5. Am I speeding up a working process or am I speeding up a broken process?