Most churches are not short on data. After years inside Planning Center, Rock RMS, or similar platforms, the average congregation is sitting on thousands of data points spanning attendance patterns, giving history, volunteer engagement, and guest follow-up. What they are short on is clarity. The gap between having data and knowing what to do with it has been one of the quiet frustrations of church operations for years.
Shelby Black is trying to close that gap. She is co-founder of Bellwether, an AI-powered tool that lets ministry leaders have a plain-language conversation with their own church data, and she leads growth and partnerships at Clever, a digital agency serving churches and nonprofits. Her background spans sales, marketing, ministry, and social impact work, with 13 years spent in or alongside the local church. She is not building Bellwether as a software product. She is building it as a pastoral care tool.
We sat down with Shelby to talk about why traditional reporting fails ministry leaders, what it looks like when a church finally asks the right questions about its own people, and where the line is between useful insight and overreach.
Q. Shelby, let’s start with something I think a lot of church leaders will immediately recognize. Most churches are sitting on years of attendance, giving, volunteer, and engagement history inside Planning Center or Rock RMS, but very few leaders feel like they actually understand what that data is telling them. Where do you think ministry teams get stuck?
Most ministry teams aren’t stuck because of missing data. They’re stuck because pulling a report means knowing what to ask, knowing how to phrase it technically, and then having bandwidth left over to do something with what you find. Ministry leaders were hired for their people skills, not their database skills. The problem is that untapped data is telling a story leaders can do SO much with if they can just ask easily — so that is why we created Bellwether. So we could just… ask.
The emotional piece is the part we prefer to skip over. When a leader suspects something is wrong, attendance sliding, a volunteer going quiet, a whole demographic slowly disappearing, looking for confirmation is hard. It can feel like asking for a verdict. So the easier move is to keep going and trust your gut. But your gut can’t hold 4,000 names in its head at once and people are drifting away in the meantime.
Q. You said something I want to pull on — “trust your gut.” Because most church software already produces reports. A leader can already open Planning Center and see totals and trends. So what feels broken about that traditional reporting model, and how does a conversational tool actually move someone from “what happened” to “why did it happen” and “what should I do now?”
Traditional reports answer the question you already knew to ask. Totals, trends, snapshots. All useful. But ministry rarely lives there. Ministry lives in the in-between questions: Are the people we baptized last year still here? Which small groups are actually retaining new people versus just holding their core? Who wants to get connected but isn’t in a group yet, and do we have groups with open spots?
Those questions have never fit a report template.
What changes with a conversational tool is where you start. Instead of navigating a system to find a pre-built report, a leader can just ask what they’re actually wondering. That shifts the posture from data retrieval to curiosity, and that’s usually when something useful surfaces.
I think church leaders in the next five years are going to expect this kind of access as a baseline. Not as a power-user feature, just as how the tools work.
Q. And that posture shift matters because we’re not talking about abstract numbers. We’re talking about people. Names. Stories. Absences that nobody has noticed yet. You gave this example — someone asks Bellwether “who used to be engaged and has quietly disappeared?” What do you hope that actually unlocks on the ministry side?
I hope it unlocks a phone call. Or a conversation in the lobby. That is honestly it.
My husband Andy and I started building Bellwether out of a personal experience of feeling forgotten by a church we loved. We had been deeply involved for years, but when life got busy and we missed church or paused on volunteering, no one noticed. Not because anyone was unkind. They were just busy growing the church, staff transitions happened, and we were not “loud” about our lack of connection. That is the story behind so many people who leave a church. They do not storm out. They drift.
When a pastor asks who has gone quiet and Bellwether surfaces a list of names, those are not data points. They are people who may be in a season of grief, burnout, doubt, or transition. The data does not tell you which one. But it tells you who to call. And that phone call — made by a real person who actually knows them — is the whole point. The technology just makes sure no one falls through the cracks before anyone even knew to look.
Q. That personal story is what makes this different from a software pitch. But I want to go somewhere that I know church leaders are going to go the moment they hear “AI” and “member data” in the same sentence. The privacy question. Pastors are stewards of deeply personal information — giving records, attendance patterns, family situations. What should they understand about how that data is protected when Bellwether is processing a query?
A. Pastors are right to be cautious here, and I want to applaud that instinct. The concern is legitimate. Most AI tools work by sending your data to a large external model to be processed. That means member names, giving records, and contact information are traveling outside your walls in ways you may not fully understand or control.
Bellwether is built differently. We mask personally identifiable information before any query is processed, and we do not send raw member data to third-party AI models. The system is designed so that the analytical work happens without exposing who your people actually are at the data level.
We have also made a firm decision as a company not to open pastoral care notes to AI querying. That was not a technical limitation, it was a values decision. Some information belongs to the pastoral relationship and should stay there. A church’s job is to protect its people, and a church’s technology should reflect that same commitment.
Q. And that values decision feeds directly into something I think pastors will push back on even after the privacy question is answered. Because even with the right guardrails, there’s a concern that a tool like this starts to encroach on discernment, on pastoral intuition, on the work of the Holy Spirit. Where do you think the line is between what Bellwether can surface and what only a pastor can do with it?
A. The questions that stick with me are the ones that reveal how much churches have been flying blind on things that really matter.
One question I find deeply meaningful: “Show me first-time guests from the last six months who never came back, and cross-reference with whether anyone from our follow-up team made contact.” That question used to require pulling three different reports, manually matching them, and spending an afternoon in a spreadsheet. Now it is one ask. And the answer tells a church exactly where their guest assimilation process is succeeding or breaking down.
Another area gaining traction is volunteer health. “Which volunteers have been serving consistently for over two years with no rotation, no break, and declining engagement scores?” That answers a burnout question. Churches care deeply about their volunteers but rarely have the visibility to be proactive rather than reactive about burnout. When the data can prompt that conversation before someone quietly steps away, that is pastoral care operating at a different level.
Q. Those examples, the guest who never got a follow-up call, the volunteer quietly burning out, both point to something bigger happening here. Data access used to belong to whoever knew the system, usually the executive pastor or the one person on staff who could build a report. What happens inside a church culture when that changes, when any ministry leader can just ask a direct question and get a useful answer?
A. This is one of the things I find most exciting and most underappreciated about what plain-language data access actually makes possible.
Think about a children’s ministry director who has never been able to pull her own retention numbers. Again, not because she did not care, but because she always had to route that request through someone else, wait for it, and then try to interpret a spreadsheet that was not built for her question. Or the question was too complex to be pulled by simple reports. Now she can ask directly. She can see which families came for a season and did not come back, and she can flag that for follow-up herself.
That is not just efficiency. That is ownership. When people across a staff team start owning their own data, you get better decisions made faster, quicker follow up and relationship building, and leaders who feel equipped rather than dependent.
The tension that can come with democratizing information is real, though. Transparency surfaces things that were previously invisible, and not all of those things are comfortable. Healthy church cultures will navigate that well. It just requires leadership that is committed to using insight to care better, not to manage more tightly.
Q. You mentioned that once leaders realize they can ask harder questions, the ceiling gets higher fast. What are some of the more surprising questions you are seeing churches actually bring to Bellwether — the ones that would have been genuinely painful to answer manually?
A. The questions that stick with me are the ones that reveal how much churches have been flying blind on things that really matter.
One question I find deeply meaningful: “Show me first-time guests from the last six months who never came back, and cross-reference with whether anyone from our follow-up team made contact.” That question used to require pulling three different reports, manually matching them, and spending an afternoon in a spreadsheet. Now it is one ask. And the answer tells a church exactly where their guest assimilation process is succeeding or breaking down.
Another area gaining traction is volunteer health. “Which volunteers have been serving consistently for over two years with no rotation, no break, and declining engagement scores?” That answers a burnout question. Churches care deeply about their volunteers but rarely have the visibility to be proactive rather than reactive about burnout. When the data can prompt that conversation before someone quietly steps away, that is pastoral care operating at a different level.
Q. Last question, Shelby. Speak directly to the pastor or executive pastor reading this. When they hear AI, data, and church systems mentioned together, what do you most want them to walk away understanding about what Bellwether is actually for?
A. I want to say something that I deeply believe. You got into ministry because you care about people. Not spreadsheets, not dashboards, not systems. People.
Bellwether exists because we believe technology should serve that calling, not complicate it. The goal was never to turn your church into a data-driven organization in the corporate sense. The goal is to make sure that the people who are quietly slipping away get noticed. That the families who showed up twice and never came back get a call. That your volunteers are cared for before they burn out. That your leaders have the visibility they need to shepherd people well at scale. No matter the size of your church.
You do not have to become a data person to use it. You just have to keep being a pastor and let the tool help you see your people more clearly.
That is what we are building. And we think it matters.
