Reducing Unwanted Messages With Machine Learning
When you send a mass text you need to pay attention to the replies to limit risk of legal action.
When you want to stop receiving unwanted messages from an automated texting service, you simply reply with “STOP”, or another pre-defined keyword, and the system will no longer send you messages. Many people don’t realize this though, and they reply with an emotional response that can be quite… colorful. These are real replies sent in response to mass texts from Call-Em-All customers, and it’s just the tip of the iceberg.
- You’re a day late and a dollar short, you suck!
- Hope you rot in hell a$$hole
- U guys are the worst
I could go on and on and on with examples. These replies are creative, often hilarious, and can be very offensive — but here’s what isn’t — a $1500 fine per message per contact. That’s what your company could be liable for when you send out unwanted messages. More on that in a minute.
My company, Call-Em-All, is a voice broadcasting and mass texting service. We’ve been in business for 12 years and we specialize in helping companies send important alerts, notifications, and reminders to large groups of contacts. What we don’t do is deliver political messages, telemarketing or promotional messages, or fundraising messages. We like to think of ourselves as the good guys in the industry and customers who don’t adhere to our strict Responsible Use Policy — get the boot. Our customers must have consent from their contacts in order to send them messages.
Whether it’s a staffing company trying to fill open jobs, a medical practice sending out immunization reminders, a trash company notifying customers about a change in service, or even an apartment complex sending out maintenance notifications — at some point, they are going to contact someone who doesn’t want to hear from them. Usually it’s because the phone number was transferred to a new person, or the contact no longer has a relationship with the organization sending the message. The bottom line is that they’re going to tell you to stop texting them, which puts the responsibility on our customers to stop.
The FCC’s Telephone Consumer Protection Act (TCPA) is a federal law that was created to curb text message spam and protect consumers. If you ignore the act, every illegal text that you send can carry a civil fine. These can be anywhere between $500-$1500 per individual text message! As consumers have become increasingly exasperated with receiving a flood of unwanted advertising on their smartphones, it has led to a rise in class actions. Large companies like American Eagle, General Electric, Coca-Cola, Burger King, General Motors, Papa John’s, Jiffy Lube, and countless others have faced lawsuits seeking huge payouts ($1-$250 million).
A small company could be financially ruined from sending one unwanted text to thousands of people without getting their consent first.
For large corporations, they could face a lawsuit where they end up paying millions just to avoid the lengthy process of litigation that could damage their brand forever. Consent isn’t really that hard to get. So, think before you send, and pay attention to your replies. It’s all about getting consent.
It’s not uncommon for a large company to send out 10,000 or even 100,000 messages at a time, and send out multiple messages to the same people. Reading every individual response and managing these replies can be a time consuming task. Not to mention, the ongoing costs of sending messages to outdated contacts.
For our customers, we decided to take a proactive approach to solving the problem. To start, the first time you send a text message to a phone number that hasn’t received any prior communication, we append the phrase “Reply STOP to stop anytime” — this is pretty common to most automated messaging services. Where the artificial intelligence comes into play is when our platform receives messages. Using our own custom built machine learning algorithm, we ‘auto-magically’ detect messages that clearly demonstrate intent to stop getting more messages. So messages like the ones above and these examples below, won’t end up getting ignored;
- Please remove me
- You have the wrong number
- I’m not interested in this anymore
The contact is then auto opted-out from receiving further messages and you can worry a less about a potential lawsuit or expensive fines.