What the Automation Fails of 2018 Can Teach Your Small Business
As automated technology continues to make life easier, it also continues to come with its fair share of hiccups. For small business owners, the risk of automation mistakes is that they can create a lasting negative impression and derail your success.
Here are a few tips for your small business to keep in mind, in light of some notable failures in 2018.
1. Thoroughly vet any third-party AI providers
Keeping customer information secure is one of the most important responsibilities entrusted to a company. If customers don’t feel safe, they will take their business elsewhere.
Unfortunately, as automation becomes an ever-larger part of the business world, security risks grow too. Every application or device connected to the Internet of Things (IoT) becomes one more potential portal to sensitive data.
This can be especially worrisome for small businesses, which typically don’t have the resources to staff internal teams to build or manage AI applications. Many turn to third-party AI vendors for resource-heavy services like chat and automation—and that means putting a lot of faith in those third-party tools to maintain security.
But, this doesn’t always work out. Some of 7.ai’s customers found out the hard way just how dangerous it can be when a portion of the chat provider’s data was exposed. Technically the breach occurred in late 2017, but users weren’t notified until well into 2018 (which is a fail in and of itself).
If you’re outsourcing automation for your small business—or if you plan to do so in the future—make sure to carefully assess providers, and regularly touch base with whichever companies you sign with to ensure their security is locked down.
And because there’s no concrete guarantee that you’ll be able to avoid a breach, you can minimize some of the risk by building a cybersecurity incident response into your business plan. IBM has a helpful list of common response pitfalls to avoid as you craft your plan.
2. Assist your voice assistants
Voice assistants can be helpful in controlling various tasks around the office, but without the proper setup, they can also cause problems.
In one scenario last year, an Alexa device accidentally sent a recorded private conversation as a message. According to Amazon, the device misheard various voice cues during the conversation, which prompted it to communicate with someone without the user even noticing.
That doesn’t mean you need to go unplug your office Google Home or Amazon Echo. But you should be careful about how you use it.
If you plan to integrate any tech with your smart assistant, look for a list of verbal commands. You can see exactly what the assistant needs access to (and limit everything else—including your contacts).
You can also make small changes like adjusting wake words, customizing commands, and taking other security measures to further prevent any mishaps.
3. Don’t over-automate
While automation can improve your small business, over-automation can decrease productivity.
Tesla saw this firsthand when it attempted to use a robot to assemble fiberglass mats—a delicate part of the company’s battery pack product. CEO Elon Musk later admitted that the process essentially turned a simple human task into a complex one, requiring more resources and resulting in more mishaps.
Find a balance in your small business by automating only the processes that increase productivity and won’t lead to total disaster if an error occurs.
When in doubt, automate only those processes with proven use cases. Many tried-and-true automation services keep a database of successes to give potential customers an idea of just how effective automation could be.
4. Understand the limits of artificial intelligence
Exercising caution with artificial intelligence will safeguard your business from the potential pitfalls of machine learning.
While some organizations are successful in using AI technology to improve customer interactions, others end up making embarrassing mistakes.
For example, AI face-recognition software in Ningbo, China, accidentally identified esteemed businesswoman Dong Mingzhu as the culprit of jaywalking when it mistook a passing bus’s signage for Dong herself. The news spread quickly, and Ningbo police had to issue an apology to Dong.
Police forces in the U.S. use another AI technology to target high-crime areas, but many claim the algorithms are biased and could lead to racial profiling.
Incidents like this can damage the credibility of your company.
One way to safely implement AI in your small business is to use it primarily for things like data analysis and marketing strategies—areas that interact more with hard data and don’t require as many soft-skill interactions.
As a final note here, if you invest everything into one platform or ecosystem, only to find out later it doesn’t do what you’d hoped—or worse, causes problems instead of fixing them—you may be stuck without the resources to reinvest.
You can reduce this risk by investing slowly and verifying AI compatibility with other systems. Amazon and Google both have thorough lists of compatible apps and devices, and most other reputable AI platforms will have similar “works with” lists that you can check for integration.
Whether you’re just starting out or have been running a small business for a while now, everyone can learn from these automation fails.
The bottom line? Automation is helpful, but it’s not infallible.