Feedback & Metrics Dashboard

Pick the few numbers that tell you how the customer side of the business is really going, put what customers do ahead of what they say, and tie every red cell to a customer comment and an owner.

Customer lifetime value rests on strong research (Gupta, Lehmann & Stuart, 2004). Survey scores are weaker: the net promoter score is weak or disputed as a growth predictor (Keiningham, Cooil, Andreassen & Aksoy, 2007; Baehre, O'Dwyer, O'Malley & Lee, 2022), and the customer effort score is practice whose claimed edge comes from its authors (Dixon, Freeman & Toman, 2010). The fixed funnel is weakly supported (Lavidge & Steiner, 1961, as reviewed by Vakratsas & Ambler, 1999). Targets and ratios on this page are rules of thumb.

Use it when:

• You report on customers or marketing every month or quarter and the report has grown into a wall of numbers nobody acts on.
• You run your own business and want to see early whether customers are staying, coming back and recommending you.
• Satisfaction looks fine but revenue or renewals feel softer than the scores suggest.

Skip it when:

• You need to understand why customers behave as they do: start with 6.12 Voice-of-Customer Notes.
• You need to work out what one customer is worth and what it costs to win one: use 5.3 Unit Economics Sheet.

A dashboard is a bet that a handful of numbers, read on a fixed rhythm, warns you sooner and more honestly than anecdotes. The metrics differ a great deal in how much they can carry. Treating customers as assets with a lifetime value is well founded and supports real choices about what to spend on winning and keeping them (Gupta, Lehmann & Stuart, 2004); the popular ratio and payback targets attached to it have no peer-reviewed source. Funnel stages such as leads and qualified leads borrow the idea of a fixed sequence of effects (Lavidge & Steiner, 1961), which later reviews found little support for. The effort score arrived with a claim of better prediction made by its own authors (Dixon, Freeman & Toman, 2010), and the net promoter score's founding claim (Reichheld, 2003) has not held up in independent tests.

So the page puts behavior first: retention, repeat purchase, churn and revenue record what customers did. Survey and funnel numbers come second, as early but softer signals. The comments section is there because it explains the numbers you can trust.

Write the decisions this dashboard should inform. A metric that serves none of them belongs in the Optional library, not on the page.

Strong answer: "Decides where next quarter's retention budget goes and whether we keep the Saturday opening hours. Owner: head of customer service. Reviewed the first Monday each month with the shop managers." Not "to track performance".

Decision 1 it informs:

Decision 2 it informs:

Owner: Reporting period:

Review rhythm and meeting:

Illustrative example: the business and all numbers are made up.

A family-owned chain of three garden centers in Sweden has a loyalty club. The customer service lead rebuilds a twenty-line report into one page.

Step 1: the dashboard decides where the spring campaign money goes and whether weekend staffing changes. Reviewed monthly with the three store managers.

Step 2: share of club members who bought in the last quarter (58 percent, target 62 because the spring mailing goes to lapsed members), repeat purchase within 90 days (31 percent, target 35) and revenue per club member.

Step 3: satisfaction after a store visit, split by weekday and weekend.

Step 4: theme "queues on Saturdays", quote "I gave up and left the trolley", explaining the weekend drop. Theme "staff know plants", quote "she saved my hedge", explaining strong weekday repeat buying.

Step 5: start with weekend repeat purchase. A second checkout on Saturdays for six weeks, owned by one store manager, reviewed in March.

The page changes one thing at once: the next meeting begins with the queue, not with the satisfaction average, which had looked fine.