All lagging, no leading
A record of the past, no view of the future
Most MI packs are 90% lagging indicators and 10% leading. The pack confirms what already happened — SL last week, AHT last month, attrition last quarter. The audience reads it, files it, and starts the same conversation the following week. Leading indicators — hiring funnel health, training pipeline, adherence trend, customer-effort score from this morning’s contacts — are usually missing or buried, even though they would let the operation act before the lagging outcome lands.
The result is an MI function that documents history accurately and influences the future barely at all.
Why operations under-invest in leading indicators
Three reasons recur.
Leading indicators are harder to construct. A lagging metric is “what happened.” A leading metric is a measurement of something that predicts what will happen, which requires the producer to have a causal hypothesis and to have validated that the indicator actually predicts. The work is more thoughtful and the producer has to defend a position.
The audience hasn’t been trained to read them. Operations leadership is comfortable with SL, AHT, CSAT. Leading indicators — hiring funnel conversion, adherence-trend slope, customer-effort score — require explanation. The producer who introduces them has to do the educational work alongside the reporting work.
The credit for leading indicators is delayed. A lagging metric that’s reported well produces a quick “good report.” A leading metric that flagged a problem six weeks before it landed produces a quick “we were warned” only if someone tracks the warning back to the indicator. Operations rarely do this consistently, so the producer of leading indicators isn’t obviously rewarded for their effort.
The six leading indicators worth adding to most MI packs
Hiring funnel health. Applications received, offers made, acceptances, starts, training pass rate. A drop in any of these predicts a headcount shortfall 2–3 months out. Most operations track end-state headcount and discover the gap after it’s landed.
Training pipeline. New-starter cohorts in pipeline, expected go-live dates, the ramp-up curve assumption. Production capacity in three months is largely determined by these numbers today.
Adherence trend (rolling 4 weeks). A persistent drift in adherence reliably precedes an SL impact two to four weeks later. Tracked as a trend rather than a single week, this is one of the cleanest leading indicators available.
Customer-effort or customer-emotion sample. Sample-based measurement of how hard customers found the contact (from speech analytics or a short post-contact survey) reliably leads CSAT by 4–8 weeks. The CSAT trend is downstream of the effort trend.
Repeat-contact rate. Customers contacting again within X days of a previous contact. Rising repeat rate predicts FCR drop, complaint volume, and CSAT decline.
Forecast bias (cumulative). Cumulative over/under-forecast tracked over time. Persistent bias predicts schedule miss-fit which predicts SL impact. The tracking-signal diagnostic catches this if anyone’s looking.
How to introduce them without disrupting the pack
The structural move is to add a small “leading indicators” section to the headline page, no more than four metrics, each shown as a trend line rather than a single number. The audience starts seeing them weekly; over a quarter they internalise what each one means; over a year they start asking about them when they don’t see them.
The pace matters. Introducing six leading indicators at once overwhelms the audience and the section gets ignored. Introducing two, getting the audience comfortable with reading them, then adding two more, then two more, produces a pack that quietly transforms over a year. Operations that try to make the change in one go usually find the new section gets skipped.
The conversation that changes the audience’s expectations
The single most useful conversation is with the operations director, in private, before the new indicators land in the pack. “I want to start tracking these two things. They predict the metric you already care about. I’ll show them in the headline section. The first time one of them flags an issue, I’ll bring you the diagnosis and the recommended action.”
The conversation does three things. It briefs the executive on what the new metric means before they have to read it. It earns sponsorship for the change. And it sets the expectation that the leading indicator will produce an action when it flags, which is the only way the discipline survives.
The honest counter-case
Some leading indicators turn out to be poor predictors. The hypothesis was wrong, the correlation was spurious, the indicator drifted with operational changes. The discipline is to test each leading indicator over a quarter or two and retire the ones that don’t earn their place. The pack that’s full of unvalidated “leading” indicators is no better than the pack that’s full of unhelpful lagging ones. The validation work is part of the discipline.
Conclusion
Most MI packs report the past accurately and predict the future barely at all. The fix is to add a small set of validated leading indicators, introduce them at a pace the audience can absorb, and brief leadership on what they mean before they have to interpret them. Operations that build this discipline turn their MI function from a record-keeper into an early-warning system. Operations that don’t continue producing accurate reports of outcomes they could have prevented.
Next in the series: The MI nobody acts on.
Pair this with leading vs lagging indicators, causal chains in MI, and designing meaningful MI.