Kaseya CEO: Why AI adoption is below industry expectations

Business data is fragmented and change management is hard

Interview Adoption of generative AI for enterprise customers isn't taking off in the manner many in the industry expected – and there are major obstacles in the way, according to Rania Succar, recently appointed CEO at Kaseya.

The bots and the tools are not very good...

"I'm very bullish on the future of AI to reinvent enterprises and small businesses and unleash productivity and growth," the exec tells The Register at the recent DattoCon event that saw 1,500 people congregate in Dublin.

"That said, the adoption has been much slower than most people expected," Succar adds. Uptake is "nascent in enterprise" and "very nascent in SMBs."

One reason is "the data is not connected; AI is only powerful when the data is connected. So if you want an agent to handle customer success, that agent needs to not only go into your customer success platform and tooling, but it also [needs] your CRM, your inventory system, and your financial systems."

There are some enterprise solutions trying to do this, however the "bots and the tools are not very good," Succar says.

A recent benchmark built by academics indicated LLM-based agents performed poorly on standard CRM tests.

"We need to build the solutions around connecting data. I see this as an opportunity for Kaseya and MSPs in the future to do this for small businesses, to find a way to take, you know, data that's sitting in SaaS applications today. Businesses use well over 15 applications to run their business.

"Customer data is fragmented, order data is fragmented, and financial data is fragmented. And if you can bring it all together through a through software... and the connected data layer, that's problem number one, then you can start to build agents on top of it that are useful.

"The agents we see are not that helpful because they can't access unified data," Succar adds.

Gartner estimates almost one of three proof-of-concepts for generative AI will be abandoned by the end of 2025. This is due to "poor data quality, inadequate risk controls, escalating costs, or unclear business value," the consultancy said in July 2024. By September, distinguished VP analysts Mary Mesaglio said at Gartner’s Symposium event in Australia:

"It is really easy to waste money on generative AI… 500 to 1,000 percent errors of AI cost estimates are possible."

Calculating business impact and return on investment are also weighing heavy on CIOs that are overseeing gen-AI projects, the analyst said.

According to Succar, "change management" – the technical and human elements of a project that can help a business switch to a new way of working – is another sticking point holding back adoption.

"It is very hard for people to find time to implement new tools. I absolutely experience this firsthand. We're all way over subscribed with our day to day schedule, and there's no time to learn something new. So we need to work on the change management side of it, too… that's why we see the high abandoned rates you're talking about."

Corporate governance is also certainly a "big issue," she tells us, but cites an example of a customer that created interaction-based rules around the how the AI agent could respond to the customer.

Customers need to work on the structure of the data and classification so that users only have access to certain types of data.

Kaseya has woven Cooper AI into its own 365 Platform and says it is seeing promising results among managed service providers that use it, such as saving 40 percent labor time by automating repetitive manual tasks and helping users improve tech utilization of its products and services, which include remote monitoring/ management, process automation, IT service desk, threat monitoring, and network performance monitoring.

The company generates "80 percent" of its revenues via MSPs - which mostly sell to small businesses - and so wants more of them to embrace the Kaseya 365 platform.

"We are seeing there's a very uneven, as you said, adoption across MSPs, which means our job is to make it even by ensuring that we measure the degree to which our customers are adopting automation and we build the tools to get them there."

She adds: "Kaseya needs to invest in driving adoption, helping people through change management. It's important for us as Kaseya to get AI adopted internally.

"We have to create… concrete change management programs, but we also need to do it for our customers. And some of it comes from software, building software in a very intuitive way, to get the engineer at the MSP side to adopt the AI. But some of it has to do with training that we could do with our MSPs."

All of this is "very early AI," said Succar, or "pre-agentic AI," and "I think there's more we could push the envelope on, and we'd like to accelerate our AI roadmap.

"I want to be careful on what we share, so we have concrete delivery dates."

Succar is just weeks in the role, having replaced Fred Voccola who served as CEO at Kaseya for around a decade and is current Vice Chairman at the company. If she mimics the length of tenure of her predecessor, then AI should be in a more mature state when he departs.

It took two decades for the cloud to mature and yet an estimated 55 percent of SMBs now manage their workloads in the public cloud. ®