Membership Data Audit

Executive summary

A member reported never receiving the newsletter. Checking that one complaint surfaced a structural problem in the membership database that predates anyone currently on the Board.

The Society has 1,779 member records. The last newsletter reached 887 of them. Delivery itself is healthy — 98.5% of what was sent arrived. The Society is not being blocked by spam filters. It is sending to about half its own membership because the other half cannot be mailed, is not on the list, or has opted out.

Where the missing half went

Three defects account for the gap. They overlap — a record can be both nameless and undeliverable — so these do not sum to 892, and no combined figure is claimed here.

1. Placeholder addresses — 339 records (19% of the membership)

These records carry an address the system invented, in the form 31308@tucsonjazzsociety.com. They are not real mailboxes and never were. Mail sent to them cannot arrive, and no amount of list management will change that — the real addresses have to be re-collected from paper, spreadsheets, or the members themselves.

2. Nameless records — 206 records

These members have an account, a membership class, and often a real email, but no name. In the admin they display as "Member." They cannot be found by searching a name — only by email address. A volunteer looking someone up by name concludes the person isn't a member.

3. Unsubscribed — 635 addresses

These sit on the platform's suppression list and are excluded from every send by design. Some are genuine opt-outs and must stay that way. The number is worth reviewing only to confirm it reflects real member choices rather than accidental bulk suppression.

The nameless records came from two bad imports

This is the most actionable finding in the report, because the cause is specific and dated.

2 May 2019 shows the same signature at larger scale: that is the original migration onto the current website platform. The same mistake, seven years earlier, and nobody caught it then either.

29 of the 57 April names can be recovered immediately by matching each record's email address against the membership spreadsheet already in the Society's possession — including Nizar Sukkar, Kathi Sanders, Tim Ouimette, Ellie Trope, Stan Lindenfeld, Calvin Hill, Karin Aldridge, Glenn Lane, Rikki Teeter and Angela Thomatis. The remaining 28 carry placeholder or blank addresses and will need a human to identify them.

Membership standing

Separately from the mail problem, the directory's own account-status field shows the following. Each status is labelled, not colour-coded alone.

The complaint that started this

A member and former Ambassador reported never receiving the newsletter. That report is what prompted the audit, and the Society was right to take it seriously. The send logs, however, tell a more specific story, and the Board should see it before anyone acts on the complaint.

How to reconcile this with the complaint — before anyone is told they're wrong

The logs prove mail was delivered to that address. They do not prove the member reads that mailbox. Three explanations fit the evidence, and the Society should establish which before treating the complaint as mistaken:

1. The mail goes to an address the member no longer monitors. The most common explanation, and entirely consistent with the logs.
2. The mail is filtered into spam or promotions at the receiving end, so it technically arrives but is never seen. This is demonstrably happening to record #31294.
3. The record's address is not the one the member believes we hold. Because the record has no name on it, nobody has ever been able to verify it by looking it up.

Recommended handling: confirm with the member which mailbox they actually use, rather than opening with the delivery log. The log answers "did we send it" — it does not answer "did they see it," and leading with it would put a valued Ambassador on the defensive over what is, at root, our own record-keeping failure.

The membership list the Society already owns

A membership spreadsheet covering 695 unique people was provided during this review. Matching it against the live directory, person by person:

What the Board is being asked to decide

1. Authorise the April 2026 name repair. 29 of the 57 nameless April records can be restored from the Society's own spreadsheet. Clerical, no cost, no risk. Asking for: approval to proceed.
2. Assign an owner to the 339 placeholder addresses. These members cannot be contacted by email at all until someone re-collects real addresses. This needs a named person and a deadline, not a volunteer's spare evening. Asking for: an owner.
3. Decide how to re-enrol the 52 current members who have no record. They are paid members receiving nothing. Asking for: approval to add them, and a decision on whether to apologise for the gap in service.
4. Adopt an import rule. Every future bulk import is checked for blank names and placeholder addresses before it is committed. This failure has now happened twice, seven years apart, for the same reason. Asking for: adoption as standing practice.

Method & confidence

Every figure above was read directly from the live TJS administration console between 19 and 22 September 2026, in read-only fashion. No member record was created, edited or deleted. All 1,779 member records were enumerated individually rather than sampled. Counts of nameless records, placeholder addresses and account status are exact. The 695-person spreadsheet match used exact email comparison first and surname comparison second; the "possibly in the directory" group is explicitly unresolved and is not counted as either present or absent. Delivery, open and click events are as recorded by the platform's own logging.

One known limitation: membership expiration dates are not exposed in the directory's member list and could not be verified independently. Where expiration is referenced, it comes from the Society's spreadsheet and is labelled as such.

Prepared by Coach Lucy Howell, in-kind Coaching Ambassador to The Tucson Jazz Society, as part of the 100-day engagement. Internal document — not for public distribution.