Predictive Intelligence

They assumed it was an agency-wide response time problem. It was one zip code.

Summit Valley Home Care was losing 8 to 10 referrals every month. Leadership had a general sense that response time was the issue. Dazie showed them exactly which referrals were lost, where they were lost, and why. The answer was not what anyone expected.

SV
Summit Valley Home Care
James Boateng, Chief Operations Officer
Predictive Intelligence

Eight to ten referrals lost every month. No one could explain the pattern.

Summit Valley Home Care had been tracking referral fallout as a monthly total for two years. The number sat on a report. Leadership looked at it, acknowledged it was too high, discussed response time improvement strategies, and moved on. The number did not improve.

The problem was the framing. Looking at total referrals lost per month is not the same as understanding where they were lost, when, from which sources, for which reasons, and whether the pattern was predictable. James Boateng, COO, recognized they were managing a symptom without understanding the disease.

The revenue impact was real. At an average of $3,200 per admission, losing eight referrals per month was costing the agency more than $300,000 in annual revenue. Every month of delayed diagnosis extended that cost. Leadership was aware of the loss but not in a position to address it without knowing its actual cause.

From a single monthly total to a live, root-cause breakdown with dollar values attached.

Before Dazie
  • Referral fallout tracked only as a total monthly number
  • No breakdown by source, geography, or root cause category
  • Revenue impact was estimated, not calculated
  • Leadership assumed the problem was agency-wide response time
  • No predictive capability to identify declining conversion before month-end
After Dazie
  • Referral fallout broken down by cause: staffing, response time, payer issues, unknown
  • Geographic concentration of losses identified within the first month
  • Revenue at risk calculated in exact dollars, not estimates
  • Predictive risk scores flagged declining conversion before month-end reporting
  • Leadership addressed root cause within six weeks of identification

Predictive scoring surfaces what month-end reports cannot see until it is already too late.

When James activated the Referral Leakage Intelligence and Predictive Intelligence modules, the first thing that changed was the granularity of the data. Instead of total losses per month, Dazie tracked intake response time by source, fallout category by volume, conversion rate trajectory, and monthly revenue at risk in real dollar terms.

The predictive scoring system went further. It calculated six operational risk scores from live data, including referral conversion trajectory, and flagged declining performance before it reached critical levels. This meant leadership could see a trend building two to three weeks before a monthly report would have surfaced it.

What the data showed was not what leadership expected. Seventy percent of referral losses were concentrated in a single geographic zone. The root cause was not agency-wide response time. It was a staffing gap in one specific zip code that created coverage delays for a subset of referral sources. Other areas of the agency were performing well. The problem was localized.

What Made the Difference
The difference between awareness and action is specificity. Every leader at Summit Valley knew referrals were being lost. Dazie told them which ones, where, and why. That shift from general awareness to specific intelligence made a targeted fix possible. A general fix aimed at agency-wide response time would not have solved a geographic staffing problem.

Live referral intelligence with predictive risk scoring built in.

Referral-to-admit conversion rate by referral source
Average intake response time in minutes
Fallout breakdown by root cause: staffing, response time, payer issues, other
Monthly revenue at risk from referral leakage in exact dollar terms
Predictive referral conversion trajectory flagging declining performance early
Six composite operational risk scores updated from live data

Root cause identified in month one. Fix implemented in week six.

70%
Of referral losses traced to a single geographic zone, not agency-wide response time
1 ZIP
Specific staffing gap identified as the primary driver of referral loss
6 weeks
From root cause identification to geographic coverage gap addressed
$300k+
Estimated annual revenue at risk from referral leakage, now quantified and presented to ownership
"We were losing 8 to 10 referrals per month and could not explain the pattern. The Dazie intake module helped us see that 70% of losses were happening because of one staffing gap in one zip code."
JB
James Boateng
Chief Operations Officer, Summit Valley Home Care

General awareness of a problem is not the same as understanding it.

Summit Valley Home Care had been aware of their referral loss problem for two years. They discussed it in leadership meetings. They tried improving agency-wide response time training. Nothing changed because the solution was aimed at the wrong cause.

Predictive intelligence does not replace operational judgment. It informs it. James Boateng and his leadership team made the decision to address the geographic staffing gap. Dazie made it possible for them to know that was the right decision to make.

The lesson from Summit Valley is not that predictive tools find problems leaders cannot see. It is that without the right level of specificity, operational awareness becomes operational paralysis. Leaders know something is wrong. They do not know what to do about it. Specificity is what converts awareness into action.

All Case Studies View Plans →