When records are disconnected, community needs become harder to understand.
This tool was developed by Damiyr to explore how fragmented data and disconnected systems can distort an organization’s understanding of whom it serves—and how those blind spots can affect decisions about community resources.
Start with one person
Imagine Alex visits two food pantries during one fictional month. Each pantry keeps its own records.
Pantry A
Two visits recorded
One record for Alex
Pantry B
One visit recorded
One record for Alex
- Visits
- 3
- Records
- 2
- Person
- 1
All three numbers describe the same month. But only one answers “How many people?” If the pantries cannot connect Alex’s records, adding their counts could count Alex twice. Confusing activity with reach can weaken decisions about community resources.
Now imagine a whole network of pantries. Explore what happens when their records overlap or leave information out.
See how the count changes
These made-up records describe a group of food pantries during one reporting period. Change the settings to see why counting records is different from counting people.
Slide the slider under “People recorded at both pantries” in the toolbar below to change how often someone has a record at each pantry. Compare “Separate records” with “Records after matching.”
What the records show
Made-up records · one reporting period
A new feature can carry an old mistake further
A dashboard, automatic report, or AI assistant depends on the information beneath it. Adding a capability does not repair records that are missing, repeated, or misunderstood.
Follow Alex’s two records through a fictional organization. These are possible consequences of leaving the underlying problems unresolved.
The records start apart
Two pantries each record Alex. Without a dependable way to connect those records, the organization may count two people instead of one.
What needs attention: A consistent way to identify people, collect essential details, and correct mistakes.
A dashboard makes the count look settled
A report labels the two records “people served.” A clear chart can make an uncertain count look authoritative. Leaders may believe they have reached more people than the records can support.
What needs attention: Agree on what “people served” means, check repeated records, and show what remains unknown.
Automation repeats the mistake
An automatic report sends the same count to more teams. Staff build plans around it. Fixing the original records later also means tracing and correcting the reports that used them.
What needs attention: Check the data before reports go out, name someone responsible for errors, and track which reports use each count.
AI can turn the mistake into advice
An AI assistant given that report may repeat the overstated reach in a funding narrative or planning answer. If people act on it without checking, resources could be directed using an incomplete picture of the community.
What needs attention: Give the system checked information, require answers to show their sources, and have people review consequential recommendations.
Neglected foundations create work—and risk
Teams spend time explaining conflicting totals, repairing reports, and rebuilding trust. Meanwhile, the analysis that could identify gaps in service gets delayed. Buying another tool can add another place for the same problem to appear.
The response is practical: improve collection, agree on definitions, check quality, and give staff time to maintain the system. These tasks are part of delivering the capability.
One useful first step is making sure everyone means the same thing when they use a number.
Make the number trustworthy
Before comparing numbers, agree on what each number counts.
A reliable number needs a clear meaning, a time period, and someone responsible for checking it. Write these down before building reports or adding new features.
Click “Visits,” then click “Records after matching.” The current example reports 316 visits and 117 records after matching. The descriptions explain why these numbers answer different questions.
Visits
Each time someone receives service. The same person can visit several times.
See how this number is defined and checked
- What is counted
- Recorded visits
- Calculation
- Add the visits attached to personal records
- What is left out
- People included only in totals appear separately.
- Who checks it
- Program manager (illustrative)
- Reporting period
- Fictional period shared with the lab
Changing the definition changes the answer.
In my food banking tenure, I have found that documenting how information should be organized, setting expectations for keeping it up to date, and prioritizing requests by their importance help teams work from a shared understanding. I commissioned an accessible dictionary and assigned people to keep its definitions clear. Former colleagues later reported that it helped resolve definition disputes.