There’s likely every chance you’ve heard enough talk about AI over the past two years to last a lifetime. From threats of existential doom from naysayers to evangelists touting its potential to revolutionize everyday tasks, there is no sector of the workforce which doesn’t have an opinion on AI.
That includes AI, itself.
Automation has been a hallmark of the workplace since the Industrial Revolution itself. In theory, it should remove the drudgery of everyday menial tasks and speed up the process of production. And in practice, it has largely done just that.
But in the advent of the digital age, something has definitively shifted wherein automation became the chief sales focus of private industries, with Software-as-a-Service (SaaS) being a perfect example. Some twenty years ago, SaaS was a highly specialized niche occupied by no more than a handful of start ups whose numbers could be charitably counted on both hands. But in 2024?
SaaS has exploded into an estimated 17,000 companies serving over 60 billion customers worldwide, giving rise to sub-industries in an already subdivided industry. It’s an industry that is firmly cemented in both small-medium businesses as well as large enterprise infrastructure, all to the tune of over $261 Billion in 2022.
But in any industry, disruption is necessary for evolution. Retail had eCommerce. And hospitality gave way to independent third-party providers such as Airbnb. But what if that disruption didn’t just change an industry, but shifted its focus altogether?
That’s just what could happen with Service-as-Software (SaS.) And if you think that AI isn’t at the forefront of that shift, you may want to think again.
Why SaS Matters
SaS isn’t so much an industry in and of itself, but a mindset.
In a traditional SaaS model, software-as-a-service is simply that: cloud-hosted software that can be rented and accessed through web applications. It’s an end product for users. And while it can be customized according to specifications, it’s frequently limited in both application and functionality (to speak nothing of strict licensing terms.)
At the same time, it can be the very heartbeat of business infrastructure.
Service-as-Software, however, takes that concept one step further: by turning it around on its head altogether. By recognizing the critical role that software plays in both SMB and enterprise entities, an SaS approach means positioning a whole host of white glove services to position itself as that same heartbeat in a line of business.
That approach includes operational tasks typically not associated with an IT infrastructure: marketing, sales and strategic analysis, accounting, forecasting and vendor relations being just a very few. Yet those tasks are fundamental to any business entity—so much so, that the slightest vulnerability in any of them can affect the very lifespan of a company.
Automation helps streamline those same tasks by helping companies deliver verifiable and replicable results, making services more efficient and scalable while boosting productivity. It’s not a replacement for human talent, but an enhancement with the potential to revolutionize and transform both the public and private sectors.
Yet with any shift in industry, there are benefits and drawbacks to consider.
Sales: An AI Transformation?
To date, most of the discussion centering around the use of AI as a sales tool has centered around its potential in both training and performing sales tasks autonomously.
As any sales rep will tell you, analyzing and forecasting sales requires an overwhelming amount of time and attention spent on repetitive minutiae and details. An amount of time and attention that can effectively hamper growth and productivity.
Nor is repetition limited to forecast and analysis. Whether it’s sourcing contracts, scheduling demos, sifting through spreadsheets or following up with clients, daily tasks can bog down even the most robust of sales teams (and some of us can still remember sales and marketing prior to the digital transformation—hardly enviable times, unless you’re feeling nostalgic for the Rolodex.)
AI may have the potential to effectively handle many of these tasks through predictive analytics, helping to create a more detailed, user focused CRM which can enable reps to focus more on understanding client needs and adapt to real-time marketplace changes.
But AI (particularly large language models, or LLMs) are only as good as the data it’s trained on. And the more limited the sales data, the more limited the quality of AI predictions.
Interpreting that data correctly and putting it to use accurately requires a native understanding of an existing market, its strengths and limitations as well as how both can affect the daily workflow of a sales and marketing team.
Predictive AI models don’t necessarily provide that understanding. The most they can do is provide relevant, real-time data—and even that data is subject to a wildly variable margin of error.
AI: A Reliable Model?
Let’s move away from some of the more outrageous and highly publicized hallucinations AI models have become notorious for and take a look at the underlying reality: namely, is AI a reliable source?
That depends on what exactly you’re using it for.
One of the chief selling points behind AI is its ability to process and analyze data at a remarkable speed, collating raw information into a coherent and understandable form in a matter of seconds. In comparison, the amount of time spent by sales and marketing into researching and understanding raw data could take weeks to decipher.
But at the sake of repeating ourselves, AI is only as good as the data it's trained upon. And while the sheer amount of raw data AI is trained upon may seem impressive, the accuracy of that data is less so.
A study conducted in 2020 found that participants’ ability to differentiate between a 500-word article written by AI showed a mean accuracy of 52% compared to that written by humans (just slightly better than random guessing.) A 2023 analysis of ChatGPT answers to 180 questions posed by physicians across 17 different specialties revealed an accuracy rate of 57.8%, while research from the University of Southern California found bias in up to 38.6% of results from AI generated facts.
AI models are based on probabilities. And probabilities are at the core of scientific analysis. But as anyone in sales and marketing will tell you, there is no reliable scientific analysis in measuring consumer habits. There’s merely second guessing.
SaS: Collaboration, Not Replacement
Despite the introduction of AI-powered solutions into various industries, it’s important to remember that SaS is an approach—not a complete disruption.
Much like biology, diverse models can coexist, thrive and support one another in the work sector. They’re not necessarily inimical, but complementary. AI may have the potential to increase productivity and ultimately transform the workplace, but it is not a replacement for human talent and ingenuity. It may be a powerful collaborative tool. But it’s ultimately a utility, not a substitute.
The very nature of a utility means that it can never be fully autonomous. Human input requires oversight, quality control and implementation. AI may enhance output, but its input needs confirmation and validation which can only be provided by human attention to error.
Displacement may be inevitable in any change to a model, and a failure to adapt to those changes will cause panic. But if the SaaS model isn’t in threat of extinction anytime soon, it’s largely a result of its agile flexibility. That flexibility needs to be extended to embrace emerging technologies if it hopes to avoid stagnation.
It’s not a question of when. It’s a question of how.
Fundamental industry changes occur at breakneck speed. It’s no different in eCommerce. To see how your eCommerce business can keep up, visit Color More Lines
Comments