What do you like best about AirOps?
We're piloting AirOps, so this is early hands-on, not a year in production.
AI and intelligence. This is the best part of the platform, and the reason we're in. The intelligence is the real thing, especially for SEO and content work. It scouts, it handles backlinking, and it holds context across a task better than we expected. Wire in your MCP connections and it gets sharper fast, because now it's pulling from your actual sources instead of guessing. If output quality is what you care about, this is where AirOps wins.
Integrations. The easy part. AirOps drops cleanly into an MCP setup, so connecting our tools and sources was quick, and those connections are what make the AI as good as it is. If you're already working in MCP, this will feel native.
Support and onboarding. A real strength. Onboarding was thorough enough to get us working without a lot of trial and error, and support has been quick when we needed it. For a platform with this much going on, that counts for more than it sounds.
UI and UX. Where there's the most room to improve. You get workflows, power agents, and playbooks, and all of it is capable, but figuring out how they fit together and when to reach for which takes a while. The power is all there. The work is the learning curve to reach it. Plan on leaning on that onboarding to get over the hump.
Performance. The other soft spot. We've hit lag and a few bugs in normal use. Nothing that's stopped us cold, but enough to break flow and get noticed. For a tool you'd run every day, this is what we'd most want tightened.
Pricing and ROI. Still TBD. We're in the pilot, so putting a number on the value right now wouldn't be honest. We'll know more once we've run it long enough to weigh output against cost. The signal looks good. The proof isn't in yet for us specifically.
Where it nets out. AirOps lives and dies on its AI, and the AI is excellent, especially once MCP is feeding it context. Support has your back while you learn it. The catch is a UX that takes real work to figure out and performance that can stutter. If you want AI quality and you'll put in the ramp, it's worth a serious look. We'll report back on ROI once the pilot has legs. Review collected by and hosted on G2.com.
What do you dislike about AirOps?
The platform is genuinely capable, and most of our friction so far has come from getting that capability organized, not from limits in what it can do. These read to us as early speed bumps rather than dealbreakers, and a few targeted improvements would smooth them out:
Getting oriented across workflows, power agents, and playbooks. The three building blocks are powerful on their own, and once they click the potential is obvious. The learning curve is in understanding how they relate and how information passes between them, which took us some time up front. That meant our early energy went into setup and logistics before we could get to the content creation and automation we're most excited about. A clearer mental model would shorten that ramp, whether it's a visual map of how the pieces hand off to each other, a guided "first automation" path, or a few templates showing all three working together end to end.
Smoothing out setup and connections. Routing data between tools takes a bit more technical setup than we expected. As one example, we ran into a Notion authentication issue and worked with support to get it sorted, which they handled well. The opportunity is to make more of that self-serve: tighter setup flows for common connections, clearer messages when an authorization doesn't go through, and a quick way to confirm a connection is live before running a full playbook. Small changes there would let teams get moving faster on their own.
Tightening up playbook reliability. Occasionally a playbook will stall on a long load or need a quit-and-refresh to get going again. It's a manageable workaround today, but smoothing it out would make longer automations feel more dependable. Clearer failure states instead of an open-ended spinner, and the ability to retry or resume a stalled step without starting over, would make a noticeable difference in day-to-day flow.
None of these have changed our read that the core platform is worth investing in. They're the rough edges you'd expect this early, and the capability underneath is strong enough that smoothing the setup, connection, and reliability layers would take an already-promising tool and make it something we lean on without a second thought. Review collected by and hosted on G2.com.