7545 Irvine Center Drive

Irvine, California 92618​ USA

info@datalog.ai / Tel: +1-415-741-5520

© 2019 by Datalog, Inc. / All rights reserved  

Terms of Service Privacy Policy  

  • White LinkedIn Icon
  • White Twitter Icon

3 Easy AI hacks for powerful business results

How to avoid perfection and effectively use artificial intelligence for 16X business results.

Are you a small business owner, middle-market executive, or startup founder? How's your business doing? Do you have everything in place that you need to thrive? Of course not. We can always do better!

Conventional wisdom implies that hard work leads to results, and that more effort leads to even better outcomes. We just need to try harder, right? Not quite. According to the "80:20 Rule,"[1] we should focus on the small portion of activities that yield the best results.

So what does this have to do with artificial intelligence (AI)? With AI, our own intelligence can be augmented with a machine that has learned how to do certain tasks efficiently. An effective AI can help achieve results at much greater scale than solely with increased personnel.

Here's the challenge with AI in business. "Perfect AI" costs a lot of money. And until recently, only large corporations could afford the investment for perfect insights. Besides The Fortune 1000, there are at least 30,000 medium-sized businesses in America alone. Because it takes substantial resources to create flawless intelligent software, such perfection isn't practical for these smaller enterprises. If it were possible to benefit from machine intelligence without requiring top talent, unlimited time, and the most expensive equipment, small to medium size benefits could realize a meaningful return on investment.

Is this in reach? Yes, because useful frameworks and methods are now in the public domain[2], and reasonably priced AI solutions are available from the many startups fueled by enthusiastic venture funding during these past two years. On the talent front, thousands of talented math-minded students and software engineers have begun the transition from web and mobile applications to practical AI development. These newcomers are now ready to bring your first easy and useful AI to life. Like the transition from web to mobile applications that began with business adoption of the iPhone, programmers now see the opportunities in AI and are cross-training for these new challenges at a rapidly growing pace.

Introduction to Deep Learning


I'm going to show you how to take advantage of recent advancements in AI to gain a 16X improvement in three areas of business performance: 

  • Revenue growth

  • Cost reduction

  • Customer experience improvement

REVENUE GROWTH HACK: Use the data you already have to target customers who want more of what you have to offer.

Have you been in business for three years or more? Sales transactions, shipment details, location data, and customer profiles are ready for the taking. And for many businesses, three years is a magic number for AI model training. It turns out to be just enough for an AI to learn how to predict outcomes. Older approaches such as big data analytics may eventually find insights, but only after spending quite a bit of money and time.

Here's how a simple AI hack can quickly lead to better results without a significant investment. In the past few years, researchers and advanced AI practitioners have cracked the code on customer behavior prediction. Training AI models using seemingly meaningless data such as the weather, flight schedules, time of the day, and others, can lead to amazing insights. Here's a secret: you don't have to know WHY someone buys your product to target them as a prospect, but AI may also reveal factors unseen to human eyes.

Hack #1 Gather customer transactions that you already have, to start building your customer targeting AI. Let's take 2/3 of this data to "train" our AI model using open frameworks like TensorFlow. Then we'll test the AI model's ability to predict behavior by testing it with the remaining data. If we can predict conversion more than one half of the time, we are doing better than a visit to the casino. I've seen results with more than 90% accuracy in prediction. That's 80% of the half of the business you miss in a coin toss. This is the 16X effect at work.

COST REDUCTION HACK: Identify wasteful patterns and reduce costs.

I love business process improvement. During my corporate years, I participated in numerous waste reduction exercises, earning my LEAN / Six Sigma Black Belt. We didn't have access to machine learning (a broader class of AI that includes deep learning) approaches, yet we always discovered something that could be streamlined or redesigned to improve quality with reduced cost.

Hack #2 Gain similar benefits with an easy use of AI. Find a small thing that a human does VERY well, and train an AI to do that. In your firm, you do your best to protect your computer systems. Part of that involves complex password requirements and lockouts after only a few attempts. How frustrating!

Password resetting has become an expensive service for IT helpdesks. The average cost of a single reset by a call center representative is as much as $20 (USD). If you can automate that process, the savings are easy to calculate. Further, an employee's ability to get back to work rapidly yields additional savings. 

Using natural language understanding[3], you can allow employees to send a message requesting a reset of their password. Texts like: "I'm locked out," "change my password," or even "let me in," are understood by the AI and allow it to initiate a structured and secure process to allow the employee to get back to work.

Other examples include the on-boarding of new employees, streamlined expense reporting, reduction of vendor payment errors, and more.

CUSTOMER EXPERIENCE IMPROVEMENT HACK: Add a chatbot to your website to easily increase customer loyalty.

As described in the example for an AI-powered password reset system, natural language understanding technology offers a new and fun channel for customer engagement. Facebook Messenger and other chatbot platforms continue to grow dramatically.

A 2016 study discovered that 75% of millennials would rather give up speaking than the ability to send and receive text messages[4].

As a new modality of interaction between consumers and brands, chatbots are accelerating this year. 


The 3rd Hack: Since you already have a website, and perhaps a mobile app, adding AI is EASY via chatbots.

There are inexpensive tools available to craft a customer conversation and increase time on your site. How does this happen? It takes more time to type than it does to click. A simple "see you later" delays the exit from your website content by just enough time to multiply your message's effect by 16X.

None of these AI hacks are very expensive. Compared to the millions of dollars spent today by large corporations on AI, these hacks don't have to be perfect. They just need to make small improvements that allow you to multiply your 20 to an 80.

Easy AI can multiply your results by 16X!

[1 ] In the early 1900's, Joseph M. Juran proposed that Vilfredo Pareto's observations of wealth in 19th century Italy could be used to focus business activities for maximum results. Author Richard Koch carried this idea further in his books on an "80:20 rule." He also revealed his 16X principle in a simple formula. By abandoning efforts on underperforming investments, we can achieve a return of 16 times our current results. 

[2] TensorFlow is a software framework released to the public in 2015. It expedites machine learning model creation through an abstraction of Google's Brain AI capabilities. Stanford professor Christopher Manning indicates that TensorFlow is capable of reducing the time spent optimizing neural networks by 100 times). "How Google Aims to Dominate Artificial Intelligence," by Dave Gershgorn, Popular Science, November 9, 2015.

[3] Natural language understanding (NLU) is the next development beyond natural language processing (NLP). NLU technology empowers a chatbot to understand the intent of the human conversant (that's you!), and identify the important parts of what you have said to the chatbot. An example would be "What is the current balance of my checking account 2344553?" The chatbot understands that you are asking for financial information about your bank account, and that it should match that to the account that matches that number.

[4] Research firm, OpenMarket, summarized in a 2016 report: "Though voice calls aren’t at risk of going extinct anytime soon, millennials — who make up an overwhelmingly vast percentage of the mobile phone market — prefer not to have to speak on the phone. To be more exact, they would rather text. According to a recent survey we conducted here at OpenMarket, when given the choice between being able only to text versus call on their mobile phone, a whopping 75 percent of millennials chose texting over talking." "Shoot Me a Text: Why Millennials Prefer Text over Talk," OpenMarket.com, May 5, 2016

Copyright (c) 2018, Jack C Crawford, All rights reserved

Logo for Datalog.ai | a Managed Ventures Company

a Managed Ventures Company