QueBIT works with all sorts of organizations: mid-size, enterprise, Fortune 100, and across all industries. We’ve seen a lot of successful projects, and we’ve talked to a lot of companies who struggle to find direction and get started. There are many reasons why any given company hasn’t started down the AI/ML path but by far, the most common reason is simple: inertia. You’ve been doing things the same way with the same people for so long that making change is hard. And the perceived barriers to change – to starting an AI/ML project—are so overwhelming, that it is far easier (and maybe safer?) to let inertia carry you onwards, as is.
So let me lay out the facts, based on over a decade of building AI/ML programs and implementing hundreds of successful projects.
Basic machine learning use cases yield enormous value.
Plain and simple, most organizations have some simple problems that can be solved with AI/ML. And in most cases, these aren’t problems at the edge of your business, these are core, systemic issues that you touch every day. Will this customer buy from me again? What should I say to this group? Should I increase prices or decrease, or both? How many widgets can I sell next month? These are fundamental questions that you can get better answers to using AI/ML.
Common business problems can be solved with simple AI solutions.
Recently I was invited to have a discussion with a healthcare company looking to dip their toe into the AI/ML world. They began the discussion by asking about Tensorflow, neural networks, and Python packages. I asked them what business problem they were tackling… and they hadn’t given it any thought. There is so much advanced technology and open source code available to us, promising the moon—but we have seen time and time again that a well defined use case can be solved with simple tools (regression, decision trees, or seasonal decomposition). Don’t overcomplicate it—Google and Facebook need state of the art. You just need to start.
Scale to Focus
Clearly, 2020 didn’t go as planned. Many, many of our clients spent this year reforecasting ad nauseum, and enhancing their planning and analysis capabilities. There has also been a big upsurge in combining operational and financial data together (eXtended Planning and Analysis or xP&A). All this work takes effort – and you’re not being given more resources to get it done. So now is the time to introduce AI/ML in a way that helps you work more efficiently.
I couch this as scaling up the ordinary so you can focus on the extraordinary. Scale to Focus. Let a machine learning model create your base forecast, then you spend your time making adjustments. Use a model to handle price changes at the SKU level, then layer on manual adjustments for promotions. Let a classification model prioritize which insurance claims you pursue. This isn’t magic, this is working smarter, and focusing on what matters. Scale up to focus where it counts.
There are plenty of objections to AI/ML. Is it expensive? Not compared to the status quo. Is it complicated? If you make it complicated, then yes… but 2021 is the time to embrace the advances in data science and machine learning and put it to work for you. And if you don’t know where to begin, we can help you define a roadmap and get started. Best wishes for 2021.