Part 1: Focus on the Solution
Should you build or buy? This is one of the top questions I’ve encountered working with customers and prospects as a technology sales consultant for the better part of my 15-year career, and one that I’d like to discuss here.
Of course, as an individual in a pre-sales role, our objective is to sell technology (that’s the “what”), but the “why” is far more important, which I’ll get into momentarily.
I’ll share a quick personal example for illustrative purposes. At expert.ai, we’re working with a major organization that is using our AI solution, specifically our Natural Language Understanding (NLU) technology to automate labor-intensive work for over 500 daily users. By increasing their efficiency by about 50%, they are going to be saving over $12 million per year. This is an AMAZING achievement, and it’s why I joined this organization. I want to empower more companies to get that kind of value from their Natural Language Processing and Natural Language Understanding projects.
I know it’s possible because I’ve seen it work. Yet, when you do a Google (or Bing!) search on the success rate of AI/ML projects, the first few pages of results are dominated by statistics and studies showing that only a fraction (as low as 15% predicted by a recent Gartner report) of these projects succeed in delivering the expected outcomes that the business requires, and over half don’t even make it out of the prototype stage. Why is that?
There are varying reasons for why almost 85% of AI/ML projects fail to deliver, including being:
- Unable to build the infrastructure to support a new project
- Unable to dedicate the right people resources
- Unable to accurately predict business outcomes
- Unable to deploy models into the production business workflow
- Unable to build trust within the organization that the results are good enough to rely on
With all these factors and variables, as well as AI still being relatively new as a business-enhancing solution, it seems like the odds are stacked against most businesses. That’s where the “why” I do what I do comes into the picture. I’m passionate about helping organizations use technology to solve business problems. Specifically in the realm of AI and NLP/NLU.
The Goal: Solving Business Problems
The conversation of build vs. buy is an age-old debate that is as old as commerce itself. So, when you’re evaluating an enterprise technology solution to solve a business problem, what are the most relevant factors that you need to consider?
- Can we get the technology off the shelf with Open Source (OS) tools?
- Can we do it cheaper if we build with these off-the-shelf components ourselves?
- Should I be utilizing my internal talent pool to build this?
- I don’t need all the features that I would be buying, so could I skip them and save a little money?
Usually, in my experience, it comes down to cost and ROI. Lots of conventional thinking leads to: “if I build it myself, it will be cheaper.” Well, maybe… but what is the objective of the project in the first place?
Example: “My business needs a CRM… should I build or buy?” It’s possible that dedicating some software developers to create a CRM using open-source tools is cheaper than buying Salesforce or Pega CRM, but aren’t there other considerations? What’s the business objective? Is it to generate a revenue forecast for investors across thousands of sales reps, and allow those reps to manage their business? Remember, we’re talking about a mission-critical solution that will help transform the business, and we need to ensure the following:
- The solution must deliver the expected business value
- The solution must be reliable for the end users to build adoption
- The solution must be deployed in an expected timeframe
- The solution must be well supported for bugs, upgrades, and enhancements
- The solution must be portable and not owned by a single employee or small group of employees
- The solution must deliver an ROI
If we don’t deliver these results, we will end up in the 85% category of failed AI/NLP projects that Gartner identified.
Keep in mind that this is for traditional SaaS (Software as a Service), where most organizations understand the risks, and therefore, don’t debate build vs. buy as much. So then why are AI and ML projects considered different?
Why AI and ML Projects Are Different
Is it because these are relatively new technologies? In reality, AI and ML projects have evolved from being experimental to being a crucial step a businesses must take to stay competitive in this exponentially expanding and technology-driven world. Organizations must learn to use these technology advancements to be better, faster, smarter and more efficient than the rest… otherwise, they’ll miss out on opportunities that their competitors will take. Organizations must think of these projects just like traditional software projects.
I know from speaking with clients that there are individuals and teams who are dedicated to and passionate about creating and deploying AI solutions for their business unit, and we certainly don’t want to cut them out of the equation. In fact, we want to empower them, as well as business users, to find the best solution for the problem at hand. I am not discounting whether talented individuals CAN build on their own, it’s about whether an organization trying to solve a business problem SHOULD build on their own.
So, what if we ask a different question: what if we combine buy and build and offer a solution that empowers businesses to ‘build’ on a platform that is proven to help deliver results? This is the question, and ultimately the solution, that I’ve seen create the most success for businesses.
In Part 2, I will offer the formula to solve the challenge that many enterprises struggle with on their path to success.
