A global survey of 4,381 small-to-medium business (SMB) executives published today by Amazon Web Services (AWS) finds nearly all (96%) have embraced artificial intelligence (AI) to one degree or another but cybersecurity (53%), data quality (47%), integration complexity (47%) and skill gaps (37%) all remain significant barriers to further adoption.
Additionally, 42% of respondents cited inflexible pricing models and unexpected costs, also known as “token shock,” as their number-one frustration with AI vendors. In fact, more than three quarters (78%) said they prefer private or hybrid AI architectures precisely because they enable more predictable, fixed infrastructure costs and reduce dependency on volatile pricing models involving application programming interfaces (APIs). A total of 42% report running into challenges scaling AI because of those unpredictable costs, the survey finds.
Those issues are also impacting IT strategies, with 41% favoring virtual private cloud (VPC), while 38% prefer hybrid deployments. Only 17% are fully running AI workloads on public clouds. A total of 43% report they have ended up with a multi-cloud computing environment more by happenstance than as the result of a strategy. A full 84% said integrated IT stacks are critical but only 29% have operationalized them. Well over a third (37%) said they have also become frustrated by aggressive AI sales pitches.
Nevertheless, SMBs are increasing investments in agentic AI technologies by an of average 16%, representing their single largest area of investment on a percentage basis, with 50% focused on measuring return on investment (ROI) rather than hard costs (15%). Slightly more respondents (46%) are also prioritizing profitability and operational efficiency over pure revenue growth (43%).
Collectively, those results suggest that, despite challenges, SMBs are doubling down on their AI investments even as they encounter stumbling blocks, says Ben Schreiner, head of AI and modern data strategy business development at AWS.
Less clear is to what degree AI is enabling SMBs to achieve a sustainable competitive advantage over rivals versus simply raising the table stakes required to remain relevant as the overall speed at which new capabilities are provided to customers continues to accelerate.
At the same time, no one is quite sure to what degree increased operational efficiencies might lead to reductions in payrolls. As more tasks are automated, there is a clear opportunity to do more with less. However, history suggests the nature of the tasks that will require humans with the aid of AI to address will simply evolve, says Schreiner. “There are always going to be more problems to solve,” he says.
One way or another, continuing advances in AI should make it easier for SMBs to compete more effectively against larger rivals even though the latter may have more resources to invest in AI. Like it or not, most organizations regardless of size now find themselves locked in an AI arms race.
The challenge, as always, will be finding ways to adroitly provide more value to customers in an era where it’s not the proverbial size of the dog in the fight that matters so much as the size of its AI bite.
