Green, Amber, Red: Weekly Engagement Health Signals
A color-coded signal per placement, logged once a week by the Account Manager. Simple enough to do in under a minute. Structured enough to catch drift before it becomes a churn risk.
The most dangerous client relationship problems are the gradual ones. A missed deadline here, a communication gap there. No single moment you could point to. Just a slow drift from "things are mostly fine" to "the client is already looking for alternatives." By the time the relationship is visibly damaged, you've usually missed six or seven opportunities to course-correct.
The Weekly Engagement Health Signal in TalentDesk is designed around this problem. Every active placement has a weekly update slot — three choices: Green, Amber, or Red — with a required note that explains the rating. The Account Manager fills it out once a week. That's the whole mechanism.
What the colors mean
The three status levels are deliberately strict in TalentDesk:
- Green — the placement is running smoothly. Communication is good, deliverables are on track, the client is satisfied, and there are no open concerns from either side.
- Amber — there is a concern that needs attention. This could be a communication issue, a skill gap showing up in a specific area, a client concern that was raised informally, or a deliverable that slipped. Amber doesn't mean the placement is failing — it means something needs to be addressed before it gets worse.
- Red — the placement is at risk. There is an active problem that requires immediate escalation. This might be a direct client complaint, a repeated pattern of missed expectations, or a breakdown in the working relationship between the engineer and the client team.
The note is mandatory for Amber and Red. You cannot log an Amber or Red signal without writing something down. This prevents the signal from becoming a checkbox exercise — the note is where the account manager puts the specific thing that is concerning them, and it's what makes the history useful in retrospect.
Why weekly and not per-incident
We chose a weekly cadence deliberately, and it's worth explaining why we didn't go with a per-incident model where you log something when something happens.
The problem with per-incident logging is that the incidents that matter most are often the soft ones — a slightly flat response to a delivery review, a client who seemed distracted in the last call, a pattern of questions being answered slowly. None of these are "incidents" in the conventional sense. But weekly, you're forced to ask the question: how is this placement actually going right now? The act of answering that question once a week, even when everything seems fine, is what builds the history that tells you when something is starting to shift.
A placement that's been Green for eight weeks and suddenly goes Amber is a different situation than one that's been oscillating between Green and Amber for a month. The weekly history gives you the context to interpret the current signal correctly.
The Action Center integration
TalentDesk's Action Center surfaces two categories of health-related items:
- Pending weekly health updates — placements where the current week's signal hasn't been logged yet. These show up as a count in the Action Center so the Account Manager (or a Master Admin) can see at a glance which placements are overdue for an update.
- Flagged placements — the global dashboard shows a count of placements currently rated Amber or Red. These also appear on the main dashboard's placement grid with their color indicator, so anyone who opens TalentDesk can see the operational health of the portfolio without navigating into individual deployments.
Monthly retrospectives
The weekly health signal tells you the current state. The Monthly Retrospective logger in TalentDesk's Client Account Health module gives you the structured space to step back and review the pattern. A monthly retrospective captures what went well in the period, what concerns emerged, what actions were taken, and what to watch in the coming month. This feeds directly into the account relationship — it's the data you bring to a monthly client check-in call.
Together, the weekly signal and the monthly retrospective create a continuous feedback loop on the client side of every placement. They're not a substitute for the account management relationship — they're the structure that makes it consistent and searchable.
The early warning that would have prevented a placement from drifting to a client call on a Thursday afternoon is almost always already in the weekly signal history, waiting to be acted on. The signal board forces someone to see it.