No client leaves suddenly. They go quiet first: fewer replies, meetings pushed, logins dropping - signals scattered across your systems where nobody assembles them. This employee assembles them. Every week it reads the engagement signals for every client relationship and flags the drift while it is still a conversation, not a cancellation call.
Engagement across your systems, gathered per client and compared against their own baseline. Not a survey, not a gut check - the actual signals, watched continuously.
The client whose replies slowed, whose meetings slipped, whose activity faded: flagged with the evidence while a warm check-in can still fix it. Early is the entire value.
Every flag comes with what changed and when, in plain language. You walk into the save conversation knowing the story instead of fishing for it.
The weekly read also tells you who is thriving - the accounts ready for a bigger ask, a case study, or a referral conversation. Health cuts both ways.
Your private channel opens the minute you hire. No call, no kickoff meeting, no software to install.
It studies your business and writes your Company Brief, then learns your client roster and what a healthy relationship looks like in your world.
You grant read access once, in your portal's Access tab, to the systems where engagement shows up. Reading is the whole grant.
The first health read proposes what gets measured and what counts as drift, in plain language. You approve it once, and it becomes the standard every week is judged by.
The weekly read lands in your channel: who drifted, who is thriving, and the evidence for both. The save conversations happen months earlier than they used to.
Like every Tenfold employee, this one runs one opinionated method: each client measured against their own baseline rather than a universal score, evidence attached to every flag, and early over certain - a false alarm costs a check-in, a missed drift costs the client. When the field moves, the method moves, and you do nothing to get the update. Here is the whole method.
It reads engagement signals through grants you make once, in your portal's Access tab. Reading is the whole grant: it never contacts a client and never changes a record. Credentials never appear in chat, and you can revoke them yourself at any time.
The ones your systems already carry: communication frequency, meeting patterns, activity and usage where it exists, response times. The first read proposes the exact set from what you connect, and you approve it.
Never. It watches and tells you. What to do about a drifting client - the call, the visit, the gesture - stays a human decision, made warmer by arriving early.
It baselines every relationship on its own history. The client who was always monthly is not flagged for being monthly; the weekly client who went silent for three weeks is.
A score alone is not, which is why every flag carries the evidence and the story: what changed, when, and how it compares to that client's normal. You act on the story; the score just ranks the reading order.
They show up too, as opportunities: the thriving accounts worth a bigger conversation, a testimonial ask, or a referral. Retention tools that only see danger miss half the value.
An AI employee that runs recurring inspections on the CRM portals you manage: broken automations, dirty data, drifted settings - found before your client finds them.
An AI employee that writes a per-client project digest every Monday: what shipped, what is in flight, what is blocked - so the week starts with the truth on one page.
An AI employee that drafts a short status email per active client every week - shipped, in flight, next - in your voice, for your approval, before anyone has to ask.
An AI employee that runs your feedback and NPS collection on a rhythm, processes every response, and surfaces the themes worth acting on - not just the scores.
Requesting this role moves it up the roster - requests set the order roles open for hire, and you will hear the moment it opens.
See all 51 roles