Quick answer: Run a ninety-minute communications audit every quarter with three people. Twelve questions covering volume, voice, approvals, guardrails, and what to stop. The point is not measurement — it is catching drift early, because church communications degrade slowly and invisibly rather than failing all at once.
Why quarterly
Annual reviews catch problems a year late. Monthly reviews become a meeting people start skipping in month three.
Ninety days is the sweet spot. It is long enough for a pattern to be real rather than a fluke, and short enough that correcting it is cheap.
Put the four dates on the calendar now. Audits that get scheduled "when things slow down" never happen, because things do not slow down.
Volume and balance
1. How much did we actually send?
Count it by channel. Most teams guess low by a wide margin, especially on text. Filter your calendar by element type and the real number appears in seconds.
2. Was any channel overloaded?
Look for weeks with three or more texts, or two emails in two days. Then check your unsubscribe rate against those weeks. The correlation is usually visible without any analysis.
3. Did any ministry get crowded out?
Filter by ministry. Churches routinely discover that one loud ministry took forty percent of the communication while three others got almost nothing — not through favoritism, but because they asked more.
Voice and quality
4. Does this still sound like us?
Pull ten random pieces from the quarter. Read them out loud. This is the single best test available and it takes four minutes.
What you are listening for is not errors. It is blandness — the slow slide from your church's accent toward generic nonprofit language. If it has happened, revisit your tone and voice settings rather than blaming the drafts.
5. What was our best piece, and why?
Pick one. Name specifically what made it work — a real story, a clear single action, the right timing. Then say it out loud so the team can repeat it deliberately.
6. What went out that should not have?
Ask directly. Wrong dates, a name published without permission, a tone misfire, a Scripture reference that was off. Every church has at least one per quarter, and the value is in finding where the process broke rather than who broke it.
Process
7. How long are approvals really taking?
Measure the gap between finished and approved. If routine items are sitting more than 48 hours, either the tiers are wrong or an approver is overloaded. Your dashboard makes this countable instead of anecdotal.
8. Is anyone doing work twice?
Ask everyone who publishes. Duplicate work is invisible to leadership and obvious to the people doing it — someone is retyping content from one place into another, every week, and has never mentioned it.
9. Is the load balanced?
Count open items by person. If one person holds most of them, have the staffing conversation now rather than after they leave.
Guardrails
10. Is the human review step still real?
This is the question most worth asking honestly. Review decays quietly — careful reading becomes skimming becomes clicking.
Look back at a month of approvals. If a reviewer cleared forty items with zero changes, either the drafts were exceptional or the review is no longer happening.
11. Are our AI boundaries still right?
Read your prohibited list out loud. Did anything come up this quarter that belongs on it and is not? Gray areas your team actually encountered should be resolved and added, so the list gets more precise over time.
Check that your AI policy settings still match the written document. Churches change positions, rename ministries, and update tone — the settings should follow.
12. What should we stop?
The most valuable question and the least asked.
Every church is carrying something that no longer earns its keep: a channel with forty followers, a print piece nobody picks up, a recurring email with a nine percent open rate, an approval step that adds two days and catches nothing.
Pick one thing and stop it this quarter. Teams that only ever add eventually collapse under the weight of their own history.
How to run the ninety minutes
- Before: the communications lead pulls the numbers — volume by channel, approval times, open and click rates, the ten sample pieces. Thirty minutes of prep.
- Minutes 1–25: volume and balance.
- Minutes 26–50: voice and quality, including reading pieces aloud.
- Minutes 51–70: process and guardrails.
- Minutes 71–90: decide three things — one to start, one to stop, one to fix. Write them down with owners.
Three decisions. Not twelve. A quarter with three real changes beats a quarter with a long list nobody touched.
Who should be in the room
Three people: the communications lead, one pastor, and one person who publishes regularly. Larger groups make the honest answers harder, and the honest answers are the entire point.
The mindset that makes it work
An audit is not a performance review. If it feels like one, people will present rather than report, and you will learn nothing you did not already believe.
Open by saying plainly that the goal is finding what is broken, that anything broken is a system problem, and that the best outcome is discovering something uncomfortable early.
Churches that hold four honest ninety-minute conversations a year rarely have a communications crisis. Not because they are luckier — because they keep catching things at the small stage, when fixing them costs an afternoon instead of a season.
What churches actually find
The twelve questions are easy to read past. It helps to know what they typically surface, because the same handful of findings come up across churches of very different sizes.
More texting than anyone believed. Nearly every church that counts text volume for the first time is surprised. The number is usually forty to sixty percent higher than the team's estimate, because each person only remembers their own.
One ministry dominating. Not by design. One leader plans ahead and asks early, so their content gets built well and often. Others ask late and get less. The fix is not restricting the organized leader — it is helping the others plan.
A channel that should have been closed years ago. An account with sixty followers that someone maintains weekly. A print piece that gets recycled untouched. The maintenance cost is real and the audience is not.
An approval step that catches nothing. Content routed to someone who has approved every item without a change for a year. That is not oversight; it is delay wearing oversight's clothes. Move that content to a different tier.
Duplicate work nobody mentioned. Someone is retyping content from one system into another every single week and has never raised it because they assumed it was necessary. This finding alone often pays for the audit.
Voice drift. Read ten pieces aloud and you will usually find two that could belong to any organization. Not wrong — just not yours. That is the signal to revisit your tone and voice settings.
Notice that most of these are not failures of effort. They are the natural sediment any organization accumulates when nobody stops to look. The audit exists to make looking a scheduled event rather than an accident.
Try for free — the numbers you need for an audit, already in one place.