Artificial Intelligence for Business: Focus on the Business Problem, Not the Technology
If you are in the process of developing strategies and acquiring technologies to become a data-driven business, you are probably familiar with the frustration of researching and vetting machine learning (ML) and/or artificial intelligence (AI) solutions. The level of “noise” and “hype” in the ML/AI marketplace is staggering, especially to those just starting their transformation into the world of digital business and looking for platforms, consultants, solutions, and services. There are more than 10,000 ML/AI startup companies to sift through for platforms and services, each offering some piece of the total “digital transformation pie.” In addition, there are many enterprise-level data and ML/AI platforms and huge consulting services offered by the larger companies that come with long deployment times and mega price tags.
For those of you looking for sophisticated AI-enabled solutions, the reality is that understanding solutions from multiple vendors will be time-consuming and overwhelming. Once you select a vendor for a particular solution, it will often require a multi-year effort to drive significant value because of the level of complexity in integration, deployment, and adoption by your workforce. For example, Gartner observes that Price Optimization and Management platforms can take up to 15 months to implement for a $1 billion business or division. That is a long time to wait for a return on investment in uncertain times. Typically, only the largest of companies can afford that length of time-to-value.
Gartner confirms what we know to be true: A majority of AI projects fail to meet expectations, and many AI solutions are purchased and integrated but never fully adopted by the intended users. We think there are a multitude of reasons for this, perhaps the most significant being that ML/AI platform and solution developers are often focused on solving technical problems or deficits instead of focusing on targeted solutions that directly impact sales and revenue. Another important reason for the high failure rate is the lack of adoption by the intended users. Simply purchasing a new platform for sales management, for example, does not ensure that your sales teams will use it. To avoid the dreaded “change management” component to acquiring new capabilities, we find it far more beneficial to adopt AI solutions one at a time and embed them into existing systems as much as possible—thus helping your workforce make better and faster decisions.
In Small Bites, AI Solutions Get Big Results
The most successfully adopted AI solutions do not attempt to change entire workflows in one blow or solve all of your business problems in one multi-featured platform. The solutions that bring the fastest time-to-value and best ROI focus on solving a specific business use case. Below I describe such a case.
A $800M foodservice distributor with 20,000 daily sales transactions, 50,000 customers, and over 6,000 SKUs knew they were leaving money on the table at every street sale to foodservice businesses (restaurants, delis, etc.). They could not continually recalculate and update the price of each SKU in the context of daily volatility in costs of goods and continually changing customer purchasing behaviors magnified due to the pandemic and the uncertainty in the restaurant industry.
They tested our AI-driven price optimization solution that is built specifically for foodservice distributors to deliver customer-specific product price lists to sales reps, daily. Their customers were a diverse group located across a broad geographical region, requiring different products for their businesses. The company needed pricing that would be accepted by each customer while maintaining its desired margins. We needed to provide a solution that could be plugged into their ERP system and delivered to their existing workflow.
After testing our price optimization solution on a subset of products and customers, the company experienced uplifts in sales of 3% and in gross profits of 4%—during the pandemic They have since rolled out the solution to include all products and customers—with similar results.
Analytics2Go (A2Go) builds many AI-driven solutions like the one described above and makes sure that the solutions can work on their own as well as when combined with other A2Go solutions. This way, smaller and medium-sized businesses can build their AI capabilities one step at a time and get value at each step.
‘AI as a Service’ Offers the Whole Pie
We believe the best path for businesses to benefit from AI capabilities is to identify and prioritize their business problems that AI solutions can actually solve. From there, they must do an audit of their existing software, department-specific platforms, ERPs, CRMs, and more to decide what pieces of the pie are missing based on what is needed. Do you really need a new pricing platform or sales management platform? Or do you simply need a specific problem solved in your pricing strategy? Prepackaged platforms more often than not come with features that go unused, are unnecessary for your business, or are not a good fit and require customization. In other words, jumping into the platform mania can run the risk of ending up with only some pieces of the digital transformation pie—leaving you hungry for more.
Becoming a data-driven business requires technical capabilities better handled by third-party providers with the expertise to vet and use the newest technologies as they emerge. This is an arduous, time-consuming process that we handle as part of our service. It is far more economical to acquire cloud-based AI capabilities “as a service” and save yourselves from the high price tags and long time-to-value of the “enterprise” approach. We at Analytics2Go are technology experts who continually explore and test the latest technology, so our customers don’t have to. Don’t bite off more than you can chew!