While not the disruption many had anticipated, robo-advisors are being embraced by affluent Americans of all ages.

According to a study of investors with more than $100,000 in net worth by Chicago-based Spectrem Group, the average “wealthy” robo-advisor user is 48 years old.

While robo-advisor users are still younger than non-users, who averaged 62 years old, the age diversity of digital advice adopters eclipsed that of non-users. While 58 percent of non-adopters were over the age of 62, just 22 percent of robo-advisor users were from the same age group. Just 9 percent of non-adopters were under the age of 45.

Thus it makes sense that retirees made up a larger portion of those who did not use a robo-advisor, 43 percent, than of those who were using robo-advisors, 21 percent. Robo-advisor users also tended to be more aggressive investors than non-adopters.

“Our research consistently shows that robo-advisors are becoming increasingly accepted by wealthy investors,” said Spectrem President George H. Walper Jr. “To remain competitive, traditional advisors should lead with their unique expertise in establishing a financial plan and emphasize their ability to evaluate and react to world events.”

Among the robo-users, almost half were directed to use digital advice by a human financial advisor. Just 13 percent reported using a robo as their primary advisor. While 41 percent of robo-users were self-directed investors, a larger portion, 43 percent, described themselves “event-driven”—likely to seek professional advice for specific needs like saving for retirement or asset allocation.

Almost a third of the robo-using respondents, 30 percent, thought that robo-advisors did a better job than human advisors at picking stocks to meet risk tolerance, with another 27 percent responding that the job could be done equally well by a human advisor or a robo-advisor. Thirty-two percent felt that robos were as good as human advisors at selecting retirement plan investments, while 28 percent thought robos were better.

On the other hand, almost two-thirds of non-adopters (64 percent) felt that human advisors could do a better job selecting retirement plan investments, while 62 percent thought humans could do better at picking stocks to meet their risk tolerance.

When asked whether a human or a robo would do a better job of establishing a financial plan, most robo users, 66 percent, felt that a human advisor would be superior while 79 percent of those who didn’t use robos felt a human advisor would be superior.

Almost three in five robo-advisor users, 56 percent, did not have an advisor before signing up for automated advice, including 64 percent of robo-advisor users over the age of 61. Spectrem says this might indicate that affluent, older first-time advice seekers are increasingly turning to robo-advisors rather than traditional advice.

The most popular robo-advisor was Fidelity Go, which was used by 27 percent of the respondents; followed by Vanguard Personal Advisor Services, used by 16 percent; Schwab Intelligent Portfolios, used by 13 percent; and AssetBuilder, used by 8 percent. The overwhelming majority of robo-users, 91 percent, rated their digital advice experience as at least “satisfactory.”

For its report, Spectrem Group did an online survey of 700 investors representative of affluent people in the U.S. The survey was performed in the fall of 2016. The respondents were required to have $100,000 in net worth, not including their primary residence.