Ozan D.

The "spreadsheet to manage the spreadsheets" problem is painfully real in enterprise. Mutual action plans especially — most reps either skip them entirely or build one too late, after the deal is already sideways. Having the structure ready before you need it means you actually use it when it matters. Nice one for sharing this.

Your point about new-account vs existing-account intelligence being fundamentally different is underappreciated. With new accounts, you're pattern-matching external signals to guess readiness. With existing accounts, you already have first-party behavioral data but most teams treat it the same way — dump it in the CRM and hope someone notices. The teams I've seen do this well separate the "discover" workflow (external signals + ICP fit) from the "act" workflow (internal signals + relationship context). Different cadence, different tools, different skills from the rep.

haven't used Unify specifically, but the shift you're describing — moving from inbound/PPC to ABM for enterprise — is one I've seen a lot of platform companies work through. a few things that have worked: start with your best existing customers and reverse-engineer their profile (what signals appeared before they bought?), and lean into your implementation/consulting relationships as a warm ABM channel — those consultants already know your product and the types of problems it solves. for the outbound piece, the biggest unlock is usually going from generic platform messaging to hyper-specific use cases per vertical. curious whether your ICP is narrow enough to hand-build target lists or broad enough that you need tooling to find them at scale — that's usually what decides whether a tool like Unify pays off.

your point about backend agentic AI doing the heavy lifting is the most interesting to me though. that might be the actual path for the skeptical majority — make AI invisible so they get the benefits without the learning curve.

the trickle-down from top performers is real — but what I keep seeing is most orgs don't know *why* their top reps get more out of AI. it usually comes down to something subtle: better prompting instincts, knowing when to trust the output vs. when to sanity-check it, integrating it into their actual workflow instead of using it as a standalone thing. the rest of the team sees the results but can't replicate the process.

spot on about quota being the ultimate filter. if the tool doesn't clearly connect to hitting their number, it's just another tab they have to learn.

One thing worth flagging — the account itself matters less than whether the messages coming from it feel real. I have seen teams with multiple rented accounts sending near-identical sequences and getting flagged faster than a single account with genuinely personalized outreach. The bottleneck moved from 'how many accounts can we run' to 'can each message actually reference something specific about the prospect.'

this is a great write-up. the session isolation approach makes sense for anything with side effects, but curious how you handle context sharing between agents when one discovers something mid-run that changes the plan for another. at 24-30 concurrent sessions that coordination overhead feels like the hardest part to get right.

The coaching angle is where this gets tricky. The reps who trigger the drop don't know they're doing it — they think they're being thorough. The only reliable catch I've found is tagging the exact timestamp where questions stop being open-ended and answers shrink to under five words. When you show someone their own call going from 'tell me more about that' to 'makes sense,' the awareness clicks faster than any framework.