The metric you track is downstream of the metric that matters
The visible outcome and the invisible cause
Most MI tracks the visible outcome — service level, AHT, CSAT, first-contact resolution. Each is a real measure of operational performance. Each is also downstream of behaviours, decisions, and disciplines that the MI rarely captures. The SL number is a function of the schedule design, the adherence discipline, the routing rules, and agent-state management. The AHT number is a function of knowledge content, system performance, the call-driver mix, and coaching effectiveness. Tracking SL without tracking those upstream drivers means the operation can see the problem but not what’s causing it. The conversation in the steering meeting becomes “SL is down 4 points” followed by speculation about why.
Three downstream-vs-upstream pairings worth tracking
SL ← adherence + schedule fit + agent state. A 4-point SL drop is the symptom. The cause is in one of those three layers. An MI pack that reports SL alongside the three upstream metrics makes the diagnosis fast; an MI pack that reports SL alone makes it a guessing game.
AHT ← knowledge content + system performance + call-driver mix. AHT drifting up by 12 seconds across a month is real and operationally significant. The driver is usually one of three places — the knowledge base is out of date so agents are searching, a system change has added friction, or the contact reasons that drive higher AHT have grown in share. Each requires a different response. Tracking AHT without tracking the upstream drivers means the steering meeting has the wrong conversation.
CSAT ← FCR + AHT + emotion handling. CSAT is a composite outcome of the contact experience. The fix is almost never “train agents on CSAT.” The fix is usually in one of the upstream behaviours that produce the CSAT outcome. The MI that tracks CSAT alongside FCR, AHT, and a measure of emotional handling (often from speech analytics or sample-based QA) produces actionable conversations; the MI that tracks CSAT alone produces vague exhortations.
How to add upstream metrics without bloating the pack
The objection is reasonable: if every downstream metric needs three upstream ones to be useful, the pack triples in size. The fix is to use the upstream metrics differently — not as standalone trends, but as causal layers shown only when the downstream metric is moving meaningfully.
The structure: report the downstream metric on the headline page (SL, AHT, CSAT). On the same page, an indicator showing which upstream driver is the likely culprit when the downstream is off. Detail of each upstream driver in a follow-up section the audience reaches only when they need to diagnose. The headline pack doesn’t bloat; the diagnostic depth is available when required.
The causal chain that turns a steering meeting concrete
The most useful single addition to most MI packs is the causal chain: a structured note that ties each meaningful metric movement to a named upstream driver. “SL down 4 points this week. Driver: adherence has dropped 2 points on the Tuesday morning shift since the new starter group joined. Mitigation: coaching focus on the new starters, expected to recover within two weeks.”
The note is short. It says what happened, why, and what’s being done. The steering meeting goes from speculation to decision. Operations leadership stops asking the planning team to investigate and starts agreeing the actions. This is the single most credibility-building MI improvement available to a planning function, and the most under-used.
Why operations resist this
Three reasons.
Upstream metrics are harder to measure. Adherence is straightforward; system-performance impact on AHT is harder; emotional handling is hardest. The MI function gravitates to what’s easy to report.
The causal chain requires judgement. Saying “the driver is adherence” is a hypothesis that could be wrong. Analysts are uncomfortable putting hypotheses in writing. The MI function gravitates to what’s defensible.
The upstream metrics implicate other functions. Saying “the system-performance issue is driving AHT” is a comment on IT’s work. The MI function gravitates to internal metrics rather than cross-functional ones.
Each of these is a reasonable concern, and each is the wrong reason to stop. The hard work is the work that earns the function its credibility.
Conclusion
Tracking the downstream metric without tracking its upstream drivers means the operation can see the problem but not its cause. The fix is to add the causal layer to the MI — not as more metrics on the page, but as a structured causal chain that turns each meaningful movement into a diagnosis and an action. Operations that build this discipline find their MI moves from data to improvement; operations that don’t continue producing accurate reports of outcomes nobody can change.
Next in the series: All lagging, no leading.
Pair this with causal chains in MI, leading vs lagging indicators, and composite metrics hide truth.