Lately, there has been a lot of noise about a “Quiet Revival”: a supposedly grassroots movement of young people, particularly young men, returning to faith. But as a data strategist, I always ask: what are we missing in the sampling? If you focus solely on who is walking through the front door, you risk ignoring the quiet Exodus out of the back.

Today, we need to talk about the “Quiet Withdrawal.” It is the data-backed reality that women, specifically Gen Z women, are leaving the Church in unprecedented numbers. If religious institutions don’t address the systemic bias in their data and their culture, are they just losing members, or are they deliberately contracting their membership to focus on the default male? Let’s see what the data says about the “Quiet Withdrawal” of women. While we know The Bible Society report was incorrect, it may have inadvertently hit on an important point about the proportion of men increasing at the expense of women.

Is there data to back up the idea of a ‘Quiet Withdrawal’ by women?

In a recent post, Garbage In, Gospel Out, I wrote about a basic rule of data work: if your inputs are flawed, your outputs are flawed. The same rule applies when churches talk about growth. Headlines about a Quiet Revival can sound convincing if the only measure is who turns up on a Sunday. That is a vanity metric. A simple churn view asks a harder question: who is leaving, and why?

If an institution counts new arrivals but does not track exits, it can mistake movement for growth. A few visible joiners can mask a larger, quieter withdrawal. This is sampling bias in plain sight. If your data collection mainly captures the most engaged people, you miss those who stop giving, stop volunteering, and stop attending. They do not make a fuss. They just go.

A Data Science approach: What does the data say?

In data science, we look for anomalies and outliers. We also look for shifts in the baseline that signal a fundamental change in behaviour. For decades, the baseline was clear and visible in the pews: women were more likely than men to be religious. They attended more, volunteered more, and stayed longer.

Recent US survey work suggests that baseline has shifted for Gen Z. The Survey Center on American Life reported in April 2024 that women make up 54% of Gen Z adults who have left their formative religion. In the same piece, nearly two-thirds (65%) of young women said they do not believe churches treat men and women equally. We know that there are debates over women in Church roles.

Like statistics, people can choose whatever Bible verses suits their particular need. From a professional perspective, I was advised by one male Christian accountant that he regarded it as ‘unChristian’ to take advice from a woman (i.e. me). I got support from leaders in the organisation and myself and his team did the work in spite of him; he simply made himself irrelevant and outdated, which is one of the worst career moves you can make. He’d set aside scriptural exhortations to work hard in favour of teachings about women; a Biblical version of cherry-picking verses – data – rather than looking at the bigger picture.

This is a small anecdotal example of the sort of thinking that led to the “Quiet Revival” report in the first place; cherry-picking findings and not asking hard enough questions. It is clear that there were plenty of people questioning the data, and they are owed an apology. Even now, The Bible Society are doubling-down by looking at anecdotal and other evidence to support their view.  My concern is that this episode could make future research efforts worse as they desperately try to recover.

Organisations measure what they care about

The quote “You cannot manage what you cannot measure,” is often attributed to Peter Drucker or W. Edwards Deming, but it is a simplification of a subtle bias that people bring to data. Organisations measure, and manage, what they care about. If they aren’t measuring the “Quiet Withdrawal” of women, it is because they do not particularly care about it and it is in their interests to do so? Why would this be the case? Let’s look at the bigger ‘why’, which is also a part of data science.

This “Quiet Withdrawal” needs to be investigated, but it will only be researched if the Churches care about it. It is a trend that has accelerated into 2025 and early 2026, so it isn’t just a minor statistical wobble. The research indicates that 65% of young women feel that churches do not treat men and women equally. In the UK, a 2025 National Secular Society report on misogyny in religious charities highlighted that these aren’t just “feelings” and they validate the lived experiences match up to subtle or obvious institutional exclusion.

Those figures do not prove what is happening in every denomination in every country. They do show a pattern that any UK institution should take seriously, because it points to a specific risk. The “Quiet Revival” focused on a visible cohort of young men while losing a less visible one (often young women). If that happens, the organisation contracts and becomes more uniform, but perhaps this is intentional as safeguarding issues remain unresolved. Women will not attend places where they don’t feel safe, but that could be intentional. When 65% of a demographic tells you they feel unwelcome, that’s not an “outlier.”

Is the ‘Quiet Revival’ masking the ‘Quiet Withdrawal’?

You might have seen the headlines about the “Quiet Revival.” There is a narrative that Gen Z men are seeking out the structure and tradition of the Church. From a business intelligence perspective, this is a fascinating study in “vanity metrics.” If the Church of England or the Bible Society only measures “new arrivals,” the “Quiet Revival” looks like a success. But if you apply a basic churn analysis, the picture changes. If you gain 10 young men but lose 12 young women, you aren’t growing; you are contracting and becoming more homogenous.

Is the noise about the ‘Quiet Revival’ masking the ‘Quiet Withdrawal’?

The “Quiet Revival” is effectively masking the “Quiet Withdrawal.” This is a classic sampling bias. If your data collection methods primarily target the loudest, most active participants, you will miss the “quiet” ones who are simply stopping their direct debits and slipping out the door during the final hymn. Anglican Vicars will not call the Police if there is an abuse risk; they will call their Diocese instead, and the two together will act as Judge, Jury and Executioner in dismissing the abuse through forgiveness. It is not in their interests to protect women from abusers, but it is in their interests to protect the reputation of their Churches. They will not get past women’s biology or sex to see the ‘spirit-soul’, and that signals danger to women as we defer and disappear. As I’ve discussed in my AI and Data Strategy work, focusing on the “wrong” growth metrics is a recipe for long-term failure. A “revival” that alienates 50% of the population isn’t a revival. I’d argue that the “Quiet Revival” is a demographic pivot toward a smaller, more insulated base of the default male. It will be easier for the Church to promote ideas about the “forgotten men” than to admit that it is demonstrably against their Christian beliefs to improve safeguarding, and that a Vicar will never pick up the phone to call Police immediately.

The Silence between the Data.

The silences are also important; the silences between the data. Some Churches were minimal in their congratulations to the new female Archbishop of Canterbury, while giving a lot of online real estate to a local male Bishop. The silences, and the ‘minimal viable product’ congratulations, say it all. The absence of prayers for safeguarding victims, the silence, says it all. Sometimes the data just isn’t there, and it is not measured because people do not care about it.

Why are women leaving? The ‘Trad Wife’ narrative and systemic misogyny

Data doesn’t exist in a vacuum; it is driven by cultural factors. When we look at the qualitative data: the “why” behind the “what”: the drivers of the Quiet Withdrawal become clear.

1. The ‘Trad Wife’ Trap: There is a growing online narrative, often echoed in more conservative church circles, that glamorizes the “traditional housewife” role. While there is nothing wrong with choosing a domestic life, the data suggests that the aggressive promotion of this narrative alienates professional, ambitious women. In 2026, women are leaders, scientists, and CEOs. When the only model of “godly womanhood” offered is one from the 1950s, women with 21st-century careers simply opt out.
2. Persistent Safeguarding Issues: We cannot ignore the data on institutional trust. The Church of England has been rocked by persistent safeguarding scandals. A 2026 Church Times article on online misogyny highlighted how social media has become a breeding ground for attacking women who speak up about these issues. Organizations like WATCH (Women and the Church) have been sounding the alarm for years, but the data shows that the institutional response is often seen as “too little, too late.”
3. The Equality Gap: As mentioned, 65% of young women don’t see equality in the pews. If a woman can lead a department in a FTSE 100 company but isn’t allowed to lead a small group or have a say in church governance, the cognitive dissonance becomes too great to ignore.

“In data science, we say ‘Garbage In, Gospel Out.’ If your datasets ignore half the population’s dissatisfaction, your strategy is effectively – and even wilfully – blind.”

What should the Bible Society and the Church of England do next?

If I were consulting for the Church of England on their data strategy, I wouldn’t tell them to start a new marketing campaign. I would tell them to start a “Data Due Diligence” project. The “Gospel Out” will only be as good as the “Data In.” Here is how they can fix the strategy:

1. Move Beyond Vanity Metrics

Counting “bums on seats” is the religious equivalent of counting “website clicks” without looking at the bounce rate. The Church needs to measure engagement and churn. They need to know why people are leaving, not just how many are staying. This requires sentiment analysis and honest, anonymous feedback loops. Example: A Church of England Church gave a paper survey to Church attendees to ask them about Church culture in the Parish. This is a perfect example of sampling bias, because only people inside the Church would get a copy of the paper questionnaire. I have no doubt that they gave themselves A+ as they “mark their own homework” by being very selective about whom they gave the homework to!

2. Create Auditable Safeguarding Proof

In the world of AI governance, we talk about independence, transparency and auditability. The Church needs the same, and the Church of England have been repeatedly advised that they need to bring their safeguarding up to secular standards. Safeguarding shouldn’t be a “black box” process. There needs to be auditable, transparent proof that reports are being handled correctly and that survivors are being supported. This is about building a data infrastructure that prioritises safety over reputation. The Church of England needs to be very honest on the theologically-based reasons for its unwillingness to be victim-centred in its approach to safeguarding. In 2024, The Economist noted that the 27 recommendations in the Makin report “reflect similar recommendations in dozens of previous safeguarding reports over 40-plus years that the CofE has previously chosen to ignore or disregard”. There is substance to that charge. The CofE set up an Independent Safeguarding Board in 2021, for instance, but it was beset by internal conflicts and was closed in 2023.

3. Address the ‘Quiet Withdrawal’ through Bias Mitigation

The Church needs to look at its leadership data. Who is making the decisions? Who is being promoted? If the data shows a bottleneck where women are excluded from high-level strategy, that’s a systemic bias that needs to be “debugged.” You can’t fix the “Quiet Withdrawal” if the people at the top don’t see it as a problem.

4. Implement AI Observability

Just as we use AI observability to track how models perform in the real world, religious institutions should use data to track the “health” of their communities. Are there “pockets” of misogyny that are driving people away? Are certain narratives causing a spike in disaffiliation? Data can surface these issues before they become terminal – or perhaps even track that women are successfully being subdued out of attending Church altogether.

The Data Science of Inclusion

The “Quiet Withdrawal” is a signal. In the world of Cloud Computing and Big Data, we know that if you ignore a signal for too long, your system crashes. The Church is at a crossroads. It can continue to celebrate the “Quiet Revival” of a few, or it can look at the hard data of the “Quiet Withdrawal” of the many. Inclusion is a strategic perspective. If you lose the women, you lose the community-builders, the educators, and the next generation of leaders. If they do not welcome women as leaders, then it would be more honest to say that is their belief. It is disingenuous to quote verses, such as Galatians 3:28, while doing the exact opposite and focusing on biology rather than the “spirit soul”. I help organisations navigate complex data landscapes to find the truth, even when it’s uncomfortable. Whether you are a corporate giant or a historic institution, the rules of data remain the same: if you put garbage strategy in, you will get a “garbage” future out. It’s time to stop ignoring the data and start listening to the women who are quietly walking away. The “Gospel” of your institution’s survival depends on it.

---

Want to learn more about how to use data to surface hidden trends in your organization? Check out our AI and Data Strategy services or explore my thoughts on detecting bias in data.