Just what is Artificial Intelligence ask 10 people and get 25 answers…. even from the experts. Sorry I am not the expert here. There seems to be some confusion between Automation’s and Artificial Intelligence (AI) or maybe not. To someone looking in from the outside automation’s seem to be magic. For example when a cookie auto-fills your web entries for you. Being a Technology guy I see that as a simple automation, AI not so much. Yet put enough Automation’s together and it can be magic, maybe even low end AI.
So the real life business question is AI, Automation’s Yes or No. That is a good question to ask and I will offer some ways to decide. Since most of what business can cost effectively implement is more of an Automation than AI. That is where I will focus.
A surprising number of business tasks are still done by people repeatedly doing the same work over and over again. When it comes to data I call that Brute Forcing it, or for Data Transfer using a Sneaker Network. Once I get Data in a digital format I NEVER EVER want to manually enter it EVER again. Lets look at some real life use cases.
One of my clients had a unique lending model. They offered a loans to businesses based on their monthly credit card sales. Their lending model was for the mostly automated. My part was to automate one part of the process. When a prospect had credit issues they shared the loan application with a partner who specialized in high risk credit. My job was to take that manual process and automate it. In addition the lender had a 72 hr window to send the loan out to the high risk lender then if declined respond to the prospect. I created a Time Based Workflow automation that: When a loan rated low, it triggered the automation to start running, it was sent to the high risk lender, and if it was declined or accepted before the 72 hr deadline our people were notified. If the high risk lender lapsed the 72 hr window the automation would advise the prospect that we were declining the loan also. Result no more missed deadlines, we generated more revenue by referring every weak loan not depending on associates to remember to refer the loan, and saved the time it would take a person to decline the loan.
For a Mortgage broker there are legal deadlines when processing a loan. For one who was very frustrated about missing deadline the Process Flow Builder Automation was the trick. First we laid out his process and the deadlines. Then put those time bound Automation’s into the Salesforce Process Flow. At each of the 8 steps the Broker gets a notice via email with a link directly to the loan that is due to be processed. This saved time both finding the right data and missed deadlines and resulting consumer complaints to the Gov with the resulting fines. Problem Solved…!
A charity who was a Client of mine raised millions of dollars a year by doing mailings. When the return envelope came back their team Manually entered the donors info. Ouch..! We know who the donors are we have their data why re-enter it? The solution was to use a QR code to automate loading the donors data. All that we had to do was enter the amount & check or credit card number. The result was an increase in donations due to sending a thank you letter right away. To quantify the savings measure the time spent before Scanning and the Time Spent after, the delta is then based on the salaries you would have paid.
In a large company small things matter, they add up. So if you can save a few minutes per person and there are a lot of persons doing a task…. well it adds up. Currently I am working on a project where Tech Support & Customer Service will be opening new cases for customers. We looked at how to we make this easier. Enter Workflow Auto-Fill. Once the person selects the Account from a look up & saves, the Workflow will Auto-Fill many empty fields with information from the Account. They don’t have to enter the Address, City, State etc. We did the same for the Key Contact once selected from the Account’s Contact list the Auto-Fill will do the rest. Time is saved and the data will be cleaner a win win all the way around. Cost a few hours writing code. Savings to be determined.
Conclusion AI Yes or No depends:
Automation and AI are here to stay make the most of them where they make sense. Don’t get caught up in chasing AI or Automation Shiny Objects. Where automation’s & AI don’t work is where the task is a one off that is not repeated very often. A task that is small and not repeated often is also a poor choice for automation or AI. Neither AI or Automation’s will work well with bad data, either they just won’t work or they will do things you don’t want done. Testing them a lot will help here. Do the Math start by the measuring the time workers spend doing the tasks before and estimate for after, use their salary times the hours, then the cost to create the automation – AI, that is an easy way to calculate the Return On Investment.
Not sure if any of this applies in your world on the off chance it does it might be worth a quick phone conversation to figure out what your next steps might be.