How to Know When to Automate Your Business

When is the wrong time to start automating your business? I'm Sarah, the founder of Carly AI, and we help businesses make work fun again by adding AI and automations to increase leverage and productivity. Many people come to us with the mistaken belief that automations will make their lives better. They read things online, they think everyone is doing it, and they think that if they do this they will unlock huge gains in their business. I cannot think of anything more harmful than thinking that, because without a good philosophy of why you are automating and what needs to be automated, you are better off staying manual. If you automate before you are ready, you end up with a huge cost, tons of mistakes, and it will hurt your revenue. So how do you know which bucket you are in, and whether you should automate or not? In this video I am doing something a little different. Instead of going through an automation or workflow, I am taking a step back to discuss the first question: what is the purpose of doing anything in your business, and how can automation be part of that? If you are in the place where automation will work for you, you can probably 100x what you are doing now with very few changes. But the reality is most people are not quite ready to automate, because they lack one thing, and that is what this video is about. Whatever you do for work, the purpose of your business is to make some return. In a for-profit business, which is what I will focus on today, that means a financial return. If you are the business owner or a manager of people, your job likely involves making decisions: capital allocation, hiring, products, how to deal with a crisis, how much inventory to order, what to do for marketing. Your performance is scored by the quality of those decisions. To make good decisions you need to monitor what is going on in your business. If you run a cafe, you need to know your inventory, your staffing, your forecasted sales, the probability that a staff member does not show up tomorrow and how that affects the rest of the day. To make the most money at the lowest cost tomorrow, you typically need to monitor the past. How many pieces of bread did we sell? How many people did we need to staff the week before? What was our forecast for that day, and what was the variance? The more accurate your numbers, the higher quality your decisions, and that lets you serve customers at the lowest cost to you, at the quality they expect, in a timely manner. This is a point Andy Grove, the former CEO of Intel, made in High Output Management. If you do not have confidence in the things you are monitoring, you cannot make decisions, and knowing those unreliable numbers only gives you anxiety. Unreliable numbers are just about useless. The other important point is that you need to monitor continuously, or on some fixed schedule. If you work in sales, you might ultimately be responsible for the number of prospects you convert by the end of the month. But if you look at your calendar tomorrow and there are no calls booked, the issue is a number at the very top of your funnel: you probably do not have enough qualified leads coming in. If you only monitor the end of the process, where people convert, you are not going to make quota. That is why the best salespeople always keep the top of the funnel full. They know that monitoring the funnel now predicts how many people they will close two weeks or two months from now. That brings me to the first mistake I see when people implement automations: they try to automate a broken monitoring system. They might have a CRM, a sales funnel, an ERP, or an inventory tracker, but the numbers are not actually reliable. Bringing in automation amplifies the confusion and noise they already had. Not only does it fail to increase revenue or lower cost, because there was no reliable monitoring to begin with, it becomes infinitely harder to disentangle the right information from a broken, noisy system. So when a client comes to me asking for automations and I see they do not really have a system, the first thing we do is spend the next month sitting down and fixing the current manual system. For a lot of businesses that fixes ninety percent of what was going wrong. If you were dropping leads because your salespeople were not picking up the phone and nobody called back when voicemails were left, you will be amazed at what happens when a proper system tracks calls and makes sure the people assigned to them actually call back. From there, yes, you can automate speed to lead, an AI receptionist calling back, and so on. But clients are often happiest just fixing their manual processes. The second mistake is when the company does have a good system but there is not widespread adoption. The company might have a task manager and a handful of people use it, maybe a lot of people, but not everyone who is part of the process. The whole company is not bought in, so the records are incomplete. Every employee thinks it is not that bad because they are doing their part, and they say things like, we almost have a perfect system, if only a few people would actually update the CRM or the tracker. The issue is that until those records are complete, the data is almost worthless. It is very hard to glean insight from incomplete data. It can be something as simple as a deal value that was never updated. A five million dollar deal entered as a twenty million dollar deal obscures how the business makes its revenue. In reality the business should be going after twenty million dollar deals, but because the records are incomplete it is hard to tell which companies to target, because the true deal value is not in the project manager. And that is only the first of a host of problems. Now you have to reconcile the bookkeeping and the financials with your project management software. A company like this might ask whether we can automate it: there is a lot happening on Zoom, phone, and email, and it could update the CRM or project management software automatically. The answer is that of course that is true. But if the records are incomplete right now, we should start by using AI to fill out all the records, because real data might totally change how you do business. And the fundamental issue remains that certain people will not adopt the automation software either. The eighty percent who adopt it will still have complete records, just as they did manually, and the twenty percent who do not update it will still leave you with incomplete records and bad data. When I look back at which businesses successfully automated much of their business, they fall into two categories. The first, and this is the rare one, is the business that really had it all together. They kept great records, they were organized, they were aligned, and the second they got hold of automations they skyrocketed. Frankly that was very rare, because few businesses have it all together. There is typically a way that works, that is how they make money, and everything else falls by the wayside. Getting everything together was never a priority because the way things worked was good enough. The second group of companies that scaled effectively with AI and automation were willing to dedicate time upfront to systematizing their workflows. If that sounds terrible and annoying, I do not think it was as bad for them as it sounded at the outset. A few things made it easier. First, having me or someone on my team be the point person in charge of the rollout, so the responsibility did not sit on someone's shoulders internally and the business felt like an outside vendor was coming in to fix things up. Second, the demystification of the technology once a specialist shows you how things work, even the current manual things you may not have buy-in for. Having someone sit next to you and walk you through it, click this, it really did not take that long, demystifies how this stuff works and how onerous it is. And last, the gains from systematizing and getting organized are big and immediate. The great thing about good data is that the value is immediate. Your own data is worth a billion outside consultants trying to tell you the truth about your business. It is all in the data, being able to trust it and see what it actually says. That is the truth of your business. I like to tell people that a lot of the unlocks for your business today are probably not in AI or automation technology. It is stuff that existed ten or twenty years ago, simpler stuff that AI has made cheaper to deploy for your own company: setting up a system to see important data, tracking things automatically. That would have cost tens of thousands of dollars a month in the past, and today, because of AI, it has become very cheap to deliver. The kind of professionalization that used to be available only to big companies with big budgets can now come to very small businesses, including solo practices. And then the cherry on top is the AI and automations, which are so much more useful and fun once you have the business basics down. So if you are a business owner looking to automate your business, I hope you now realize that is probably not the first step for you. We might be able to help. Feel free to book a call with my team and we would love to see if we are a fit.