FastForward #69: If enterprise AI is hard, small business AI may be harder
Hi everyone. There was no newsletter last week as I was moderating a CEO fireside at Pax8 Beyond in Salt Lake City on Monday, then flew to Vegas on Tuesday for the Zscaler Zenith conference where I recorded a video interview with their EVP of the Agentic AI Security Startup division. Both of these were paid gigs. If you need a professional moderator for your event, drop me a line at ron@fastforward.blog. If you like my newsletter, please share this week’s edition with a friend, it really helps.💌 Sign up here.
ForwardThinking 🤔
If enterprise AI is hard, small business AI may be harder
Disclosure: I was paid by Pax8 to moderate a fireside chat with CEO Scott Chasin at the company's Beyond 2026 conference, and some of his comments from that conversation are quoted in this piece. Pax8 had no input into or review of this piece before publication. My editorial guidelines apply the same regardless.
Maybe you’ve seen the Microsoft commercial where a pizza parlor owner is puzzling over whether to bring back the $1 slice. He pulls out his laptop, fires up Excel and with the help of Copilot, generates some nifty charts to show him that he would be a fool not to. The commercial closes with a line around the block as AI has made him a working class hero.
According to the World Economic Forum, there are approximately 400 million small and medium-sized businesses worldwide. The report states that they account for around 90% of all businesses, 70% of employees and 50% of global GDP.
With those kinds of numbers, it's easy to think of a small business as having similar needs to their larger counterparts, just on a smaller scale. But just as children are not small adults, neither are small businesses simply small enterprises. They have a very different set of requirements. AI doesn't always turn out like you think it will, and small businesses could be less tolerant of that risk.
In a recent personal example, I asked Claude to copy an existing invoice and create a new one with new numbers, a task any invoicing software should handle easily. Instead, it kept rendering three fillable fields for my banking information in black, even though the example I provided was clearly gray. Three tries later, it still couldn’t get the field color right. Unfortunately, in this case AI didn’t speed up my business, it slowed it down. And I run into these kinds of obstacles on a regular basis.
While SMBs are clearly flirting with AI, survey data suggests they still worry about technical obstacles. That’s why many turn to managed service providers, or MSPs, to act as their outsourced IT team and handle the underlying complexity of running the business.
MSPs are responsible for updating systems and keeping them secure. In the not-too-distant future, they’ll also be expected to build agents for clients, while helping them get more comfortable using AI — or setting things up so they don’t even realize they’re using it.
Filling the trust gap
In the end, the Microsoft commercial is more aspirational than practical for many small businesses. It presents a world where you ask a question and get the perfect answer that pushes your business into new territory. Reality is much messier. A recent Goldman Sachs small business survey found that while 76% of small businesses are already using AI, only 14% have fully integrated it into core operations, half worry about security and data privacy and 73% say they need more training and implementation support.
This tracks with my own experience. The productivity payoff that vendors like Microsoft are pitching is at best theoretical. In practice, I have learned it can be hit-or-miss, and sometimes AI simply makes the process more cumbersome.
If enterprises with all their resources are running into problems making AI work effectively, that level of difficulty is going to be compounded for small businesses. I was recently at the Pax8 Beyond Conference in Salt Lake City. Pax8 is a marketplace MSPs use to buy and resell cloud and AI tools for their small business customers.
The message for the MSP audience was clear across keynotes and the fireside chat I moderated with CEO Scott Chasin: everyone needs to be using AI. Just as you would expect at a conference like this, Chasin was delivering an optimistic vision where SMBs will begin accelerating their use of AI with the help of the MSPs, who will certainly play an important role.
"We're entering this phase where AI is now useful to every business on the planet, and I think that that's a massive opportunity. I think the awakening is starting now, and so I expect as we get into next year the demand is going to come," he told me at the fireside chat.
That's the bullish view, of course, but as my own experience shows, execution is harder than vendors want us to believe. This is especially true when it comes to SMBs where the technical hurdles tend to be magnified, imprecision and mistakes could be more costly and a lack of ROI even more impactful.
Then there's the trust issue, which looms even larger. It's going to be a struggle to get SMBs to come along after they get a few bad answers, especially when the owner's name is on the line. The old maxim, 'once burned, twice shy' is going to come into play here.
Chasin acknowledges that trust in these tools has to be earned and the answers you get verified, but he believes if you can harness it, AI can give what he calls the "greatest cognitive shell exoskeleton" that any of us can have. "It is a superpower, but you have to learn how to use it," he said. Perhaps, but that promise still comes with plenty of friction and small business owners have to navigate that ambiguity.
~Ron
What's new on the blog 📰
Salesforce returns to its acquisitive ways to help fill in agentic gaps
I've been following Salesforce for a good number of years, and when they bought Fin (formerly Intercom) this week, it got me thinking about a time when the company shut down M&A because of activist pressure.
Drew Houston has built a legacy any founder would be proud of
The story goes that Drew Houston came up with the idea for Dropbox while on a long bus ride after realizing he had forgotten his thumb drive at home and had no way to access his files.
Nineteen years later, Houston announced he would be stepping down after a transition period to launch a new AI-focussed startup outside of Dropbox, and he'll get a chance to see if lightning can strike twice.
Confluent's Jay Kreps returns to his large company roots with acquisition by IBM
I caught up with Confluent CEO Jay Kreps at IBM Think earlier this month in Boston. We chatted about what it was like being back inside a large company after being acquired by Big Blue at the end of last year.
As IBM tries to leverage acquisitions like Confluent to help orchestrate AI, capturing data in real time becomes a critical component for models using data to understand context.
"We bring the ability to do these very rich transformations for processing of that data, and the last mile in the chain is the ability to plug that into these AI models and serve up context off of that."
Aaron Levie's take on how AI moves the work goalposts — and what that means for jobs
I connected with Box CEO Aaron Levie recently on the FastForward on PPN podcast.
"What I tend to see is that the work just expands based on the kind of tool capabilities that we have, and I've seen no evidence at any kind of macro scale where that doesn't happen," Levie told me on the podcast.
As developers shift to AI coding, the nature of engineering is changing
AI is inevitably going to have a huge impact on just about every job, but perhaps none more so than the developer role where it is already being widely used.
I spoke to execs at Intuit, Amazon and MetLife to understand just how much the role is shifting and what it could mean moving forward in how we think about engineering.
"From a philosophical change, the skills that are required by the builders, the software engineers, are to really understand software architecture, to know what great software actually looks like, and how you make sure that the coding agent builds software that actually solves the problem, is high quality, can scale, and is not full of defects," Intuit CTO Alex Balazs told FastForward.
News of the Week 📣
Databricks security vision continues to develop with Panther acquisition
In March, Databricks announced it intended to get into the security business. It made sense. You track all of your other company data in Databricks, why not do the same with security info, but security is a very specific market, and simple aspiration doesn't always get you there.
That's why this week's announcement that the company intends to acquire Panther is so important. It gives some teeth to that earlier security announcement in the form of software, expertise and customers.
At the time, Sanjeev Mohan, a former Gartner analyst who now runs his own shop, SanjMo and has covered database technologies for many years, told me that Databricks was engaging in a different sales motion and it wouldn't be easy to sell to a security buyer. He believes this acquisition could fix that by addressing the concerns he had.
"Along with a well respected AI SOC platform with 100-plus prebuilt integrations, Databricks also gets a team of engineers and former SOC analysts who already speak the CISO's language and carry credibility," he told FastForward. "That takes Lakewatch from an open SIEM vision to an actual detection engine and agentic SOC workflows underneath it."
Devin Pratt, an IDC analyst who covers databases, agrees that this is a meaningful acquisition for the company and elevates the security product. "Lakewatch gave Databricks the security lakehouse foundation. Panther adds the layer the SOC depends on day to day: detection, investigation and response. This is Databricks moving to replace the legacy SIEM, not sit beside it," he said.
Panther was founded in 2018 and raised over $140 million, per Crunchbase data. The last meaningful raise was a $120 million back in 2021. And in an ironic twist, Snowflake Ventures (Databricks' biggest rival), was a Series B investor.
Automated customer service is suddenly hot as acquisitions by Salesforce, Zendesk show
Just this week, Salesforce announced it intends to acquire Fin (the artist formerly known as Intercom until recently), for $3.6 billion, and with it got a proven model, over 30,000 customers and a whole lot of expertise. In March, Zendesk announced it was bringing Forethought, a similar company, into the fold. It shows that the next AI frontier after coding could be automated customer service.
IDC analyst Oru Mohiuddin points out that both Zendesk and Salesforce are making a similar strategic bet. "Autonomous AI resolution is the future of customer service, and acquiring proven capability is faster than building it," she said.
Adrian McDermott, Zendesk's CTO, sees it similarly. He said that while the talent alone was worth it, the technology enhances what his company is trying to do with agents. "They have creatively solved a problem which is plaguing a lot of AI deployments in general in the industry, which is how do I absorb common human agent behavior patterns and other patterns, take that data and turn it into procedures and skills and map all of that to an AI agent," McDermott told FastForward in a recent interview.
Mohiuddin says the key difference between the two deals is scale. "Zendesk is building the most capable AI-native customer service platform in its category," she said. "Salesforce is building something larger — an enterprise AI platform where customer service is one layer among many, underpinned by a data foundation and workflow intelligence acquired separately. The capital deployed, the breadth of the stack, and the cross-sell opportunity all operate at a different magnitude."
The question is whether each approach is appropriate for each company, and whether customers will respond positively to automated customer service. It's one thing to automate coding, it's another to introduce it to a customer-facing environment where those customers may resent interacting with a bot.
AWS wants to help you automate tracking your cloud costs
Back in 2021 when money was still cheap and interest rates were low, companies weren't all that concerned with controlling cloud costs, but as the economy worsened in subsequent years and interest rates rose, management became more interested in finding ways to save money. Enter FinOps tools, which help set up cost guardrails and shut down orphaned projects that can add up to significant costs.
As with everything else these days, there's now an agent for that, and earlier this month, ahead of the AWS Summit in NYC, the company announced it was launching a new FinOps agent.
It's part chatbot and part automation engine, but it's still basically the same type of financial controls that you've been seeing from the cloud vendors and external software companies for a number of years now wrapped in an AI package.
The AI part is really about querying your billing information looking for anomalies, but you could turn those queries into agents that do it for you automatically without asking directly.
Chances are though that you will want a human involved to make sure it doesn't shut down a project that needs to stay up, but if it's searching for orphaned projects and other easy pickings, it saves those same humans from doing that themselves.
The tool is available for free for now in public preview, subject to monthly usage limits (like so many AI tools these days).
What I'm reading 📚
The slowtech revolution is here to kill your phone addiction and rescue your attention span
~By Amanda Silberling, TechCrunch
Microsoft Enlists AWS to Help GitHub Handle Explosive Growth in AI Development
~By James Maguire, DevOps.com
AI to cement the future of industry
~By Ingrid Lunden, Resilience Media
‘Tell Him He’s a Piece of Shit’: Meta’s New AI Unit Is a Total Mess
~By Paresh Dave & Zoë Schiffer, Wired
What I'm watching 📺
The Token Trap: How Enterprises Are Sleepwalking Into an AI Cost Disaster
~Dave is not AI, YouTube
Look who's talking 👄
"Viewed alongside the Informatica and Convergence.ai acquisitions, a coherent platform architecture is taking shape: unified data management, intelligent workflow automation, and now front-line customer resolution."
~IDC analyst Oru Mohiuddin on how this week's acquisition of Fin/Intercom fits with some other recent Salesforce purchases, as told to FastForward.
