How AI Helps a Small Church Get Its Week Back

By Wednesday my week has usually decided what it wants to be. There is a bulletin to lay out, a funeral homily to start, three emails from the vestry, a sign-up sheet nobody filled in, and a sermon still sitting as a blank page with Sunday coming whether I am ready or not. None of that is the ministry I was ordained for. All of it has to happen before I get to the ministry I was ordained for.

That arithmetic is wearing pastors down. In 2025 about one in four senior Protestant pastors said they had seriously considered leaving full-time ministry in the past year, down from roughly two in five at the 2022 peak but still a quarter of the profession standing near the door [1]. I build AI workflows for my own parish and teach other pastors to do the same at Handbuilt, and the thing I keep returning to is narrow and unglamorous. The machine is good at the parts of the week that were never the point, which is exactly where a small church needs the help.

So here is the honest version before I try to sell you anything. The time problem is real and it is measurable. AI's role in a small church is small and mostly administrative. And the payoff is time, plain hours handed back to the work only a person can do.

The Real Cost of a Pastor's Week

Before AI is worth talking about, the weight has to be named honestly. The most recent Lifeway Research survey of 1,516 pastors, run in the spring of 2025, puts numbers on the feeling most of us already carry into Monday morning [2]. It is worth reading the trend lines too, because the picture is heavy without being a crisis. Several of these figures have actually improved over the last decade.

How pastors describe their week	Share	Trend
Feel they must be on-call 24 hours a day	67%	Down from 84% in 2015
Say the role is frequently overwhelming	57%	Up from 54% in 2015, below the 2021 high
Feel ministry demands exceed their capacity	47%	Steady across the decade
Get at least one day of rest each week	78%	Down from 85% in 2015

Read those together and the shape is clear. Two out of three pastors feel they can never fully clock out. Nearly half feel the job asks more than they have to give. And the day of rest that is supposed to be non-negotiable is quietly slipping for one pastor in five. This is the soil AI gets planted in. Not a productivity fantasy, a tired person trying to protect one day off.

Where the Mismatch Actually Lives

Here is the part that reframed the whole question for me. The hours a pastor loses are rarely the hours a pastor is called to. Barna's 2026 study of pastors and AI found a consistent gap between the work pastors find most meaningful, the preaching and the discipling and the raising up of new leaders, and the work that actually fills the calendar. The meaningful work sits apart from the time-consuming work, and that gap is exactly where the tools are landing.

Reading Barna's numbers, the way pastors are currently deploying AI maps directly onto that gap.

That is the diagnosis under everything that follows. Pastors reach for these tools to clear the administrative underbrush between Monday and the sermon. The preaching stays with them. The tool is filling the low-meaning, high-hours corner of the week, which is the corner most pastors would happily hand off if they had anyone to hand it to.

What Small Churches Don't Have

Almost every guide written about churches and AI quietly assumes a staff. It imagines a communications director to hand the newsletter to, an office administrator to route the emails, a volunteer coordinator to chase the sign-up sheet. Spread across four or five people, the administrative load is real but survivable.

A solo parish does not have four or five people. It often has one, and that one is the pastor. When there is no staff to absorb it, the whole back office lands on the same desk where the sermon is supposed to get written. Here is what falls to a solo pastor with no budget for another hire:

• Nobody else drafts the bulletin, so the pastor lays out the order of service every week.
• The email from the family who visited on Sunday has no one else to answer it, so it waits, and then it is Thursday.
• Booking the room, updating the calendar, and fixing the sign-up sheet all come back to the same person.
• Nobody else writes the announcement, the newsletter blurb, or the social post for the fall series.

This is the gap the ranking articles miss. A church with a staff can spread the load across several desks, so AI mostly just shaves a little off each one. A solo pastor has no other desk to spread it to, which makes a workflow the only realistic way any of that plate gets cleared. The smaller the parish, the larger the return, because there is nobody else in the building to do this work.

What AI Actually Takes Off the Plate

When I say AI gives time back, I mean something specific and bounded. It drafts, it organizes, and it does first passes on the recurring work, and I stay the one who reviews it and sends it. Barna's task data shows pastors are already using it exactly this way [3].

What pastors hand to AI most	Share who use it this way
Brainstorming and idea generation	50%
Graphic design or visual creation	37%
Researching biblical or theological topics	36%
Admin tasks and small-group discussion questions	34%

In my own week the wins sort into three buckets.

Administrative work

The email drafting is the one I would fight to keep. I hand an agent the incoming message and the context, it returns a first draft in my voice, and I read it, fix the one thing it got wrong about the person, and send. The same pattern covers calendar and scheduling replies and the routine responses that used to sit in my inbox growing heavier. Lifeway's guidance on AI in discipleship frames this well, treating AI as an administrative assistant for policy drafting and communication rather than a stand-in for relational ministry [4].

Communications

Bulletins, the weekly newsletter, announcement copy, and the caption for a social post are all recurring writing tasks with a stable shape. An agent that knows my parish and my calendar can produce a usable first draft of any of them in the time it takes me to pour coffee. I edit for tone and truth. The draft was never going to be the finished thing, and it does not need to be.

Research and prep support

This is where the line has to stay firm. AI is a strong research partner for background reading, historical context, and the commentary sweep that used to eat an afternoon. It is not a sermon writer, and I never let it become one. It gathers so I can think. The study stays mine. Shé Langley, writing for Mennonite Church USA, puts the boundary the way I would, describing these tools as a way to reduce friction rather than to replace people, pastoral care, or discernment [5].

Why Adoption Is High But Trust Is Still Low

The strange thing about this moment is that the argument about whether to use AI is mostly over, even though almost nobody feels settled about it. Barna found that 87% of pastors use AI in some capacity, with only 13% reporting none at all [3]. That is nearly everyone, well past any early-adopter fringe.

And yet the same study found 71% of pastors feel cautious about it and 40% feel conflicted. The tools are already in the building, and the people using them are uneasy. Vendors selling speed cannot close that gap, because speed was never the real objection. The worry underneath is about integrity. A pastor afraid he is quietly outsourcing his calling needs a way of working he can stand behind, and a faster tool does nothing for that fear. It is worth taking seriously on its own terms, which is where the next section goes.

The answer to that caution is specificity, not hesitation: name exactly what gets handed off, stay the one who reviews it before it goes out under your name, and the conflicted feeling tends to settle. A pastor who can say precisely what a tool touched, and precisely what it never will, has already done most of the work of trusting it.

The Line That Doesn't Move

Knowing which parts of the week can be handed off, and which parts cannot, is the whole job. The presence in the hospital room, the discernment about a hard decision, the care of a grieving family, and the word actually preached on Sunday stay with the pastor. No workflow touches those, and no honest guide would suggest otherwise.

There is a subtler cost worth naming too. Benjamin Eggen, writing for 9Marks, argues that the efficiency of AI in sermon prep carries a hidden price, because the slow work of study forms the preacher and not only the sermon [6]. Offload the study and you may keep the output while losing the formation. I think he is right, and it is the reason my own line sits where it does. I let AI clear the desk so I have more time for the study, never less, and I only ever hand it administrative friction, the kind that has nothing to do with forming me as a preacher.

I have argued the full case for where that line belongs elsewhere. If you want the long version of the assistance-versus-authorship question, I worked through it in Is It Wrong for a Pastor to Use AI?. Here it is enough to say the line is real, it is knowable, and everything I am describing lives well on the safe side of it.

From a Chat Window to a Workflow You Own

Most pastors I talk to are already using ChatGPT, and most of them think that is the ceiling. They open a chat window, ask for help with an email, copy the answer out, and close the tab. It helps. It also resets to zero every single time, so the pastor is doing the same setup work on Monday that he did last Monday.

The next step is a workflow. Instead of re-explaining my parish to a blank chat every time, I built an agent that already knows my context and runs the same task the same way each week. The email drafting is one. A live dashboard that pulls numbers from a few scattered sources into one view is another. I built those once and now I own them, and they run whether I am inspired that morning or not.

A workflow worth keeping needs a few things a one-off chat never gets:

• Context that persists, so the tool already knows your parish, your calendar, and your voice before you ask it anything.
• A trigger that repeats on its own schedule, so you are not the one remembering to start it every week.
• A review step that stays with you, because the draft is never the send.
• A shape simple enough that you, not a vendor, can fix it when something changes.

Ownership matters more in a small church, not less. A big church can buy another monthly SaaS subscription and assign someone to maintain it. A solo pastor cannot afford a stack of tools nobody else can run, and he definitely cannot afford one that breaks the week he is out sick or forgets the login. A workflow you built and understand does not have that failure mode. It is portable, it costs nothing extra to keep, and it stays yours when the vendor changes its pricing or the volunteer who set it up moves away. Teaching pastors to build and keep exactly those workflows is the whole point of the Handbuilt Lab.

Where to Start This Week

If you take one thing from this, take this: do not try to automate your whole week. Pick the single task that repeats every seven days without fail, the one you could describe in your sleep, and hand only that to a workflow first. For most pastors it is the bulletin, the follow-up email, or the meeting notes. Get one of those running cleanly, feel the hour come back, and then add the next one.

The sorting is simple once you see it. Some tasks repeat on a schedule and ask almost nothing of your judgment. Those are where to start. Others require your presence or your discernment, and those stay with you no matter how good the tools get.

Start here (repeats weekly, low judgment)	Leave to you (needs presence or discernment)
The bulletin and order of service	The sermon's claim on the text
Follow-up emails after a first visit	The visit itself, and the hospital room
Meeting agendas and notes	Deciding what the meeting is actually for
Newsletter and announcement first drafts	The pastoral letter that has to come from you
Calendar and room scheduling	Who gets your hours this week

Start on the left. Protect the right. The whole method is learning to tell which column a task belongs in and refusing to let the two mix.

Book a Call

The promise on the front of everything I do is plain: get your week back. The Handbuilt Lab is a four-week cohort that begins the second week of September, and the first one is an invitation-only founding cohort. There is no public price and no open registration. Prospects join the waitlist and we talk first, because the conversation is how I select the cohort and how we both make sure it is a fit before anyone commits.

If your week has decided what it wants to be and you would like a say in that again, book a call. We will look at your actual week, find the one task that repeats every seven days, and figure out whether a workflow you own could carry it. That is the whole conversation, and it is a good place to start.

Frequently Asked Questions

Can AI really give a pastor back time each week?

Yes, on the recurring administrative work. The realistic gains come from handing off tasks that repeat on a schedule and need little judgment, like the bulletin, routine emails, and meeting notes. Barna's data shows pastors already lean on AI mostly for brainstorming and routine admin, which is exactly the low-meaning, high-hours corner of the week.

What AI tasks are safe for a small church with no staff to spread them across?

The same tasks a communications director or office administrator would handle if you had one: bulletin layout, newsletter and announcement drafts, first-draft email replies, and scheduling. In a solo-pastor church all of that lands on one desk, so automating the recurring pieces is the only way to clear it without hiring.

Will using AI for admin work make ministry feel less personal?

It should make it more personal, because it clears the impersonal work off your desk. The point of handing the bulletin to a workflow is to protect the hours that go to the hospital room and the study. The care, the presence, and the preaching stay entirely yours.

How much time does AI actually save, realistically?

It depends on how much recurring writing and admin your week holds, so I will not quote you a number I cannot stand behind. The honest way to find out is to take your single most repetitive weekly task, run it through a workflow for a month, and measure the hours against doing it by hand. Start with one task, not the whole week.

What should AI never be trusted with in a small church?

Pastoral care, discernment, the presence you bring to grief, and the word you preach on Sunday. Mennonite Church USA frames the boundary as reducing friction rather than replacing people or pastoral care, and 9Marks adds the warning that offloading the study can cost you the formation that study provides [5][6]. Automate the desk work freely and keep the soul work entirely your own.

Is Handbuilt a sermon-writing tool?

No. Handbuilt teaches pastors to build their own AI workflows for weekly writing and admin, with every word still in their own voice. It does not write sermons, and I would not teach it if it did. The sermon is the one thing on the list that has to come from you.

References

[1] Barna Group. "Pastors Quitting Ministry: New Barna Data Shows a Shift." Barna Group, 2026. https://www.barna.com/trends/pastors-quitting-ministry-barna-data/

[2] Aaron Earls. "Pastors Remain Committed to the Pulpit." Lifeway Research, May 29, 2025. https://research.lifeway.com/2025/05/29/pastors-remain-committed-to-the-pulpit/

[3] Barna Group (with Gloo). "Pastors Are Using AI More Than You Think." Barna Group, June 15, 2026. https://www.barna.com/research/pastors-using-ai-ministry/

[4] David Thorne. "How Can Church Leaders Use AI in Discipleship?" Lifeway Research, March 20, 2025. https://research.lifeway.com/2025/03/20/how-can-church-leaders-use-ai-in-discipleship/

[5] Shé Langley. "The top 5 AI and automation tools I recommend for churches." Mennonite Church USA, January 7, 2026. https://www.mennoniteusa.org/ai-for-churches/

[6] Benjamin Eggen. "Are AI's Promises of Productivity Worth the Risks for Pastors?" 9Marks, July 16, 2025. https://www.9marks.org/article/are-ai-s-promises-of-productivity-worth-the-risks-for-pastors/