agile# How to Use the Everyday Low Price Strategy to Create Meaningful Metrics and KPIs in Healthcare Marketplace (1/39)
A healthcare marketplace company running FDD with a team of six to fifteen people has a metrics problem. The platform connects patients with providers. It handles appointment scheduling, provider search, insurance verification, telehealth sessions, and patient reviews. The company is seven years old with 1,200 employees. The product team has eleven people running Feature Driven Development as a single feature team. (2/39)
The metrics are broken. They are vanity metrics. They look good on dashboards but do not drive decisions. A typical dashboard shows 450,000 total registered users, 120,000 monthly active users, 8,500 total appointments booked, and an average session duration of four minutes and thirty seconds. The numbers go up. They look impressive. But they do not tell the team what is working and what is not. (3/39)
Last quarter, the team built a new provider search feature using a machine learning algorithm to rank providers by relevance. The dashboard showed 68% of users adopted the feature. The team celebrated. But nobody measured whether it helped patients find the right provider faster. Nobody measured whether it reduced no shows. Nobody measured whether it improved patient satisfaction. (4/39)
That 68% was a vanity metric. It hid the truth. Patients were using the search feature but not finding the right provider. The no show rate increased by 12%. Patient satisfaction dropped by eight points. The feature looked successful on the dashboard but was actually making things worse.
The metrics need to become meaningful. (5/39)
Sam Walton built Walmart on the everyday low price strategy. The strategy was simple. He did not compete on promotions, discounts, or flashy marketing. He competed on price. Every day. Low price. Every day. The strategy required discipline, and that discipline was measurement. (6/39)
Walton measured everything. But he did not measure vanity metrics. He measured meaningful ones: cost per unit, inventory turnover, gross margin per customer, supplier lead time, and shrinkage rate. These metrics drove decisions. When cost per unit went up, he renegotiated with the supplier. When inventory turnover went down, he reduced the order quantity. When gross margin per customer dropped, he analyzed the customer segment. (7/39)
These metrics eliminated waste. Eliminating waste enabled low prices. Low prices attracted customers. The customers built Walmart. (8/39)
Walton applied the same thinking to store operations. When he opened a new store, he did not measure the number of visitors. He measured the number of purchases. Visitors was a vanity metric. Purchases was a meaningful one. The number of visitors could go up while purchases went down. The vanity metric would say the store was successful. The meaningful metric would say it was failing. Walton chose the meaningful metric. That choice saved the store. (9/39)
For a healthcare marketplace company, the problem is the same. The team measures vanity metrics. Those metrics hide the truth. Walton's strategy says: measure what drives decisions. Meaningful metrics drive decisions. Decisions eliminate waste. Eliminating waste improves the product.
## The Core Principle (10/39)
Walton's strategy was built on a simple insight. The best way to create meaningful metrics is to measure what drives decisions, not what looks good on a dashboard. He did not measure store visitors. He measured purchases. Purchases told him whether the store was working. Visitors did not. (11/39)
For a healthcare marketplace company, the situation is identical. The team measures vanity metrics that do not drive decisions. The fix is the same. Measure what drives decisions. Those metrics will improve the product.
## Four Steps to Apply the Strategy
1. Audit the Current Dashboard and Separate Vanity Metrics from Decision-Driving Metrics (12/39)
Walton audited Walmart's metrics in 1962. The audit revealed that 70% were vanity metrics. They looked good but did not drive decisions. He removed them. That created space for meaningful metrics that actually drove decisions and improved the business. (13/39)
Your team should do the same. The product manager leads a two-hour dashboard audit with all eleven team members. Every metric on the current dashboard gets evaluated with one question: Does this drive a decision? If yes, it stays. If no, it goes. (14/39)
The current dashboard has twenty metrics. Total registered users: 450,000. Does it drive a decision? No. Vanity metric. Monthly active users: 120,000. Does it drive a decision? No. Vanity metric. Total appointments booked: 8,500. No. Vanity metric. Average session duration: four minutes and thirty seconds. No. Vanity metric. Provider search usage rate: 68%. No. Vanity metric. (15/39)
Appointment no show rate: 22%. Does it drive a decision? Yes. If it goes up, the team investigates the booking flow. Decision-driving metric. Patient satisfaction score: 72 out of 100. Yes. If it drops, the team investigates the patient experience. Decision-driving metric. Insurance verification success rate: 84%. Yes. If it drops, the team investigates the verification process. Decision-driving metric. Telehealth session completion rate: 91%. Yes (16/39)
. If it drops, the team investigates the platform. Decision-driving metric. Provider response time: three minutes and forty-five seconds. Yes. If it goes up, the team investigates the notification system. Decision-driving metric. Time to first appointment: fourteen minutes and thirty seconds. Yes. If it goes up, the team investigates the scheduling algorithm. Decision-driving metric. (17/39)
Fourteen of the twenty metrics are vanity metrics. They get removed. Six decision-driving metrics stay: appointment no show rate, patient satisfaction score, insurance verification success rate, telehealth session completion rate, provider response time, and time to first appointment. The new dashboard is half the size and twice as useful. (18/39)
For an FDD team of six to fifteen, this audit should happen in one session of no more than two hours. Every metric gets evaluated with one question. For FDD, the audit should be part of domain object modeling. It is a modeling input.
2. Define One Decision for Every Metric and Write It Down Next to the Metric
Walton defined one decision for every metric at Walmart. He wrote it down. The visibility created clarity. Clarity created action. Action improved the business. (19/39)
Do the same. The product manager adds a decision statement to every metric on the new dashboard. Each statement answers one question: What decision does this drive? (20/39)
Appointment no show rate at 22%: If the no show rate exceeds 25% for two consecutive weeks, investigate the booking flow and test a reminder notification feature. Patient satisfaction score at 72: If the score drops below 70 for two consecutive weeks, run a patient feedback survey and identify the top three pain points (21/39)
. Insurance verification success rate at 84%: If the rate drops below 80% for two consecutive weeks, investigate the verification process and test a manual verification fallback. Telehealth session completion rate at 91%: If the rate drops below 88% for two consecutive weeks, investigate the telehealth platform and test a bandwidth optimization feature (22/39)
. Provider response time at three minutes and forty-five seconds: If the response time exceeds five minutes for two consecutive weeks, investigate the provider notification system and test a push notification escalation. Time to first appointment at fourteen minutes and thirty seconds: If the time exceeds twenty minutes for two consecutive weeks, investigate the scheduling algorithm and test a provider availability cache. (23/39)
These statements are written down and displayed next to the metrics on the dashboard. Last week, the no show rate hit 26%. The dashboard showed the rate and the decision statement. The team investigated and found that reminder emails were going to spam. They tested an SMS reminder feature. The no show rate dropped to 19%. The decision statement drove the decision. The decision improved the metric. (24/39)
For an FDD team, every metric should have a visible decision statement. For FDD, the statement should be part of feature design. It is a design requirement.
3. Set a Threshold for Every Metric That Triggers an Automatic Investigation
Walton set a threshold for every metric at Walmart. The threshold was a number that triggered an automatic action. This eliminated delay, reduced waste, and enabled low prices. (25/39)
Set thresholds based on the last six months of historical data. Place the threshold at the point where the metric starts to indicate a problem. (26/39)
Appointment no show rate, historical range 18% to 24%. Threshold: 25%. Patient satisfaction score, historical range 71 to 78. Threshold: 70. Insurance verification success rate, historical range 83% to 92%. Threshold: 80%. Telehealth session completion rate, historical range 90% to 96%. Threshold: 88%. Provider response time, historical range two minutes and thirty seconds to four minutes and fifteen seconds. Threshold: five minutes (27/39)
. Time to first appointment, historical range twelve to seventeen minutes. Threshold: twenty minutes.
When a metric crosses its threshold, an automatic investigation is triggered. A task gets added to the backlog and assigned to the team. (28/39)
Last month, the patient satisfaction score dropped to 69. The threshold was 70. The trigger fired. A task was added: Run a patient feedback survey and identify the top three pain points. The team ran the survey. Three pain points emerged. Patients could not find appointment confirmation emails. They could not reschedule online. They could not see insurance coverage before booking. All three went into the backlog, got prioritized, and got addressed (29/39)
. The satisfaction score recovered to 74.
For an FDD team, every metric should have a threshold based on historical data. The threshold should trigger an automatic investigation. For FDD, the threshold should be part of feature design.
4. Review the Dashboard Every Iteration and Remove Metrics That Have Not Driven a Decision in Three Iterations
Walton reviewed Walmart's metrics every week. He removed metrics that were not driving decisions. This kept the dashboard lean and focused. (30/39)
Do the same at the end of every iteration. Check each metric with one question: Has this driven a decision in the last three iterations? If yes, it stays. If no, it goes. (31/39)
At the end of one iteration, the product manager reviews the six metrics. Appointment no show rate: yes, it triggered an investigation two iterations ago that led to the SMS reminder feature. It stays. Patient satisfaction score: yes, it triggered an investigation last iteration that led to the feedback survey. It stays. Insurance verification success rate: no, it has been stable at 84% for three iterations and has not triggered anything. It gets removed (32/39)
. Telehealth session completion rate: no, stable at 91% for three iterations. Removed. Provider response time: yes, it triggered an investigation three iterations ago that led to push notification escalation. It stays. Time to first appointment: no, stable at fourteen minutes for three iterations. Removed. (33/39)
Three metrics get removed. Three stay: appointment no show rate, patient satisfaction score, and provider response time. The dashboard is lean and focused. Every remaining metric drives decisions.
For an FDD team, this review should happen at the end of every iteration. For FDD, it should be part of feature inspection.
## Closing on Decision-Driving Over Vanity (34/39)
Sam Walton did not build Walmart by counting visitors and celebrating when the number went up even as purchases went down. He built it by auditing metrics and removing the 70% that were vanity. He kept the 30% that drove decisions. He defined one decision for every metric and wrote it down. He set thresholds based on historical data that triggered automatic investigations. He reviewed the dashboard every week and removed anything that had not driven a decision recently. (35/39)
For a healthcare marketplace company running FDD with a team of six to fifteen, creating meaningful metrics requires the same approach. Audit the dashboard and separate the fourteen vanity metrics from the six decision-driving ones by asking one question: does this drive a decision? Define one decision for every metric and write it down. Set thresholds based on six months of historical data that trigger automatic investigations (36/39)
. Review the dashboard every iteration and remove anything that has not driven a decision in three iterations. (37/39)
Start by having your product manager lead a two-hour dashboard audit this iteration. Add decision statements and thresholds to the remaining metrics next iteration. Within three iterations, the company stops celebrating vanity metrics and starts driving decisions. A healthcare marketplace learned to create meaningful metrics from a retail pioneer who proved that the best way to measure success is to stop counting what looks good and start counting what drives action. (38/39)