---
title: When “I Need More Data” Becomes a Safe Place to Hide
description: One more analysis can look like diligence and work as avoidance. Learn the test that tells them apart and how to act on enough evidence.
image: https://media.babylovegrowth.ai/blog-images/organization-48833/1794548747753_Hand-drawn-evidence-and-decision-making-title-card.jpeg
---

[Skip to content](https://cognistry.co/edge/i-need-more-data#main-content)

[![Cognistry Capabilty](https://cognistry.co/hs-fs/hubfs/Logos/Cognistry%20Logo.png?width=250&height=100&name=Cognistry%20Logo.png "Cognistry Capabilty")](https://cognistry.co)

- [Home](https://cognistry.co)
- [Platform](https://cognistry.co/platform)

    - [What is Capability Engineering](https://cognistry.co/capability-engineering-system)
    - [Signal](https://cognistry.co/signal)
    - [Forge](https://cognistry.co/forge)
    - [Sim](https://cognistry.co/sim)
    - [Edge](https://cognistry.co/adoption)
- [Solutions](https://cognistry.co/overview)

    - [Customer Success](https://cognistry.co/customer-success-support)
    - [Learning & Development](https://cognistry.co/learning-development-training)
    - [Manager Enablement](https://cognistry.co/manager-enablement)
    - [Operations](https://cognistry.co/operations-teams)
    - [Product](https://cognistry.co/product-delivery)
    - [Sales Enablement](https://cognistry.co/sales-enablement)
    - [Strategy Transformation](https://cognistry.co/strategy-transformation)
    - [Talent Management](https://cognistry.co/talent-performance)
- [About](https://cognistry.co/about)
- [Resources](https://cognistry.co/resources)

    - [Get the Book!](https://cognistry.co/the-ai-lead-lambert)
    - [Recommended](https://cognistry.co/edge)
- [LOGIN](https://app.cognistry.co)

[Insights](https://cognistry.co/edge) / [Prove Value](https://cognistry.co/edge/tag/prove-value)

# When “I Need More Data” Becomes a Safe Place to Hide

![Cognistry Team](https://app.hubspot.com/settings/avatar/d41d8cd98f00b204e9800998ecf8427e)

[Cognistry Team](https://cognistry.co/edge/author/cognistry-team)

 Nov 15, 2026, 5:45:00 AM ·  22 min read

![When “I Need More Data” Becomes a Safe Place to Hide](https://media.babylovegrowth.ai/blog-images/organization-48833/1794548747753_Hand-drawn-evidence-and-decision-making-title-card.jpeg)

#### In this article

1. Loading...

#### Share this article

[Twitter](https://twitter.com/intent/tweet?url=https%3A%2F%2Fcognistry.co%2Fedge%2Fi-need-more-data&text=%3Cspan+id%3D%22hs_cos_wrapper_name%22+class%3D%22hs_cos_wrapper+hs_cos_wrapper_meta_field+hs_cos_wrapper_type_text%22+style%3D%22%22+data-hs-cos-general-type%3D%22meta_field%22+data-hs-cos-type%3D%22text%22+%3EWhen+%E2%80%9CI+Need+More+Data%E2%80%9D+Becomes+a+Safe+Place+to+Hide%3C%2Fspan%3E) [LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fcognistry.co%2Fedge%2Fi-need-more-data) [Facebook](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2Fcognistry.co%2Fedge%2Fi-need-more-data)

Copy link

#### About the Author

![Cognistry Team](https://app.hubspot.com/settings/avatar/d41d8cd98f00b204e9800998ecf8427e)

[Cognistry Team](https://cognistry.co/edge/author/cognistry-team) info@digitalcommand.co

We build systems that help organizations turn expertise into real operational capability. Our work focuses on solving a common enterprise challenge: critical knowledge exists across documents, experts, and data—but rarely becomes structured, repeatable decision capability.

[View all posts](https://cognistry.co/edge/author/cognistry-team)

The meeting ran fifty minutes. The options were clear, the owners were in the room, and then someone said it: “Before we commit, can we get one more analysis?” Everyone nodded. It was the third time. Nobody was lying, and nobody was lazy. But the decision left the room exactly as it entered.

When “I need more data” becomes a safe place to hide, the request is no longer about information. It is about exposure. It protects the person asking from being wrong in public, and the organization rewards it because it sounds diligent. The fix is not to gather less. It is to ask, before any new analysis starts, what decision it would change.

Cognistry

Ground Enablement in Business Evidence

 

Before building training, Cognistry helps enterprises diagnose what capability the work requires and whether learning is the right response.

[Visit](https://cognistry.co/)

## TL;DR

- A request for more data is justified when you can name the decision it could change. If the answer is the same either way, the data is cover, not input.
- Asking for more is a sane response to systems that punish visible error. Nobody has to be blamed for it, but the pattern has to be named.
- Define “enough” before collecting: one key question answered, a deadline, a named decision owner and a default that takes effect if nobody decides.
- Reversible decisions need less evidence than irreversible ones. Match the depth of analysis to the cost of being wrong.
- Before anyone builds a course or any other enablement play, confirm the gap is one that learning can close. That takes organizational evidence, not another dashboard.

## Table of Contents

- [Why the third request feels so reasonable](https://cognistry.co/edge/i-need-more-data#why-the-third-request-feels-so-reasonable)
- [One of five tangles: meet “More”](https://cognistry.co/edge/i-need-more-data#one-of-five-tangles-meet-more)
- [The two-sided test](https://cognistry.co/edge/i-need-more-data#the-two-sided-test)
- [When more data is the right call](https://cognistry.co/edge/i-need-more-data#when-more-data-is-the-right-call)
- [Define “enough” before you start](https://cognistry.co/edge/i-need-more-data#define-enough-before-you-start)
- [Making it safe to decide](https://cognistry.co/edge/i-need-more-data#making-it-safe-to-decide)
- [Before you build that course](https://cognistry.co/edge/i-need-more-data#before-you-build-that-course)
- [A short field guide for the next meeting](https://cognistry.co/edge/i-need-more-data#a-short-field-guide-for-the-next-meeting)
- [Promo](https://cognistry.co/edge/i-need-more-data#promo)
- [FAQ](https://cognistry.co/edge/i-need-more-data#faq)
- [Sources](https://cognistry.co/edge/i-need-more-data#sources)

## Why the third request feels so reasonable

Each request, taken alone, is defensible. One more quarter of usage data. One more round of interviews. One more benchmark. [MporgSoft’s analysis of enterprise decision-making](https://mporgsoft.com/deciding-to-decide-enterprise-data-collection-analysis-paralysis/) describes exactly this: in risk-averse organizations every new request feels individually reasonable, yet together they amount to analysis paralysis at enterprise scale. It adds that each extra approval tier gives stakeholders a way to signal diligence without committing to anything.

That is the social mechanism. Asking for more data is visible, respectable and safe. Deciding is visible, and it can be wrong. In a culture where a visible error costs a reputation, a rational person picks the option with no downside to them personally. Practitioners who write about this say the same plainly. Grant Purdy, writing on [data-driven decision-making](https://www.sufficientcertainty.com/data-driven-decision-making), argues that when the safest move becomes asking for more data and delaying the call, the behavior is career-driven rather than data-driven.

Notice what that framing leaves out: villains. The person who asks for the third analysis is usually the one who cares most about getting it right. The system taught them that being late is forgivable and being wrong is not.

![Why the third request feels so reasonable](https://media.babylovegrowth.ai/blog-images/organization-48833/1794548890280_Why-the-third-request-feels-so-reasonable.jpeg)

## One of five tangles: meet “More”

We use the word Tangles for a small set of recurring patterns that snag good decisions on the way to action. There are five: Right, More, Control, Waste and Time. They are not diagnoses and they are not personality types. They are practitioner patterns of activation friction, the ways capable people and teams get stuck between knowing what to do and doing it.

This article is about More. It sounds like this:

- “We should validate that before we move.”
- “Can we get a cleaner baseline?”
- “I’d like to see another quarter.”
- “Let’s wait for the next survey.”

None of these is wrong on its face. More becomes a tangle when the request has no stated purpose beyond comfort. The giveaway is that nobody can say what they would do differently if the number came back high, and what they would do if it came back low.

## The two-sided test

There is one question that sorts honest requests from hiding places:

**What decision would this data change, and what would we do if it came back either way?**

Run it out loud, in the meeting, with the person who made the request. Three outcomes are common.

1. **They can name both branches.** “If attrition on the team is above the baseline, we reassign the work. If it is below, we leave it.” The request is real. Set a date and collect.
2. **They can name one branch.** “If it’s bad, we’ll act. If it’s fine, we’ll probably act anyway.” The data cannot change the decision. Decide now.
3. **They cannot name either.** The request is about reassurance. That is human, and worth saying kindly: “It sounds like we want more certainty than this decision can give us.”

This is not our invention. The Conference Board’s brief [How to Make Better Decisions with Less Data](https://www.conference-board.org/research/human-capital-briefs/blg-00-1-5304) recommends holding every data request to an if-then statement: if I collect the data, how would my decision change? If the data will not change the decision, the extra information is not needed. The same brief argues that “garbage-can” decision making is curbed by thinking strategically about how data applies to decisions, rather than collecting openly and sorting it out later.

Timing matters too. Write the decision rule before the results arrive. If the team writes what counts as a yes and a no only after seeing the numbers, almost any result can be justified in hindsight. That is how analysis becomes a ritual instead of an input.

## When more data is the right call

The point is not that data is a trap. Sometimes the request is exactly right, and refusing it would be its own kind of recklessness. The National Academies’ guide on [the art and science of decision-making](https://www.nationalacademies.org/read/29265/chapter/13) lists six factors that determine how much information a decision needs:

- the experience level of the people deciding
- the quality of the information
- the importance of the decision
- how easily it can be changed
- the urgency
- the risk

The most useful of these for leaders is changeability. The guide reasons that a decision that is easy to change if wrong can be hazarded with much less information than one that cannot be undone. A pilot with an exit date, a staged rollout or a two-week trial of a new workflow does not need a six-month study first. A decision to retire a system, restructure a team or commit a budget for years does.

The same guide makes a second point worth repeating to any team: a small amount of information from a very reliable source can serve better than reams of input from armchair experts. More volume is not more certainty. Often it is more noise, and more noise is one more reason to wait.

So before approving a request, ask:

- **Is it relevant?** Does it measure the thing the decision turns on, or something nearby that is easier to count?
- **Is it reliable?** Does it come from a source the people deciding trust? Deloitte’s [workforce analytics article](https://action.deloitte.com/insight/ucyfegbtzqfrzdh4rnmy6dhn/todays-workforce-challenges-demand-data-diversity-diligence) notes that trust in data needs both confidence in its quality and trust in the people who steward it. It also warns that organizations relying on spreadsheets and ad hoc tools risk inefficient or inaccurate analyses.
- **Is it proportionate?** Does the cost and delay of getting it match the stakes of the decision?
- **Is it reversible?** If we are wrong, can we adjust in weeks rather than years?

If the answers are yes, yes, yes and yes, collect it. If they are not, the request is probably a Tangle.

## Define “enough” before you start

The most effective habit is also the simplest: decide what enough looks like before anyone starts collecting. The National Academies guide gives concrete examples. Enough might be one key question answered, or 75% of a list covered. It also advises setting a default decision that takes effect if no other decision replaces it, which lets the team gather information up to the critical point with less pressure. The guide notes that waiting for perfect information produces a default decision of inaction, which is still a decision, only one nobody owns.

MporgSoft describes the same discipline in high-functioning enterprise teams: they settle three questions before collecting anything.

1. What information would actually change the decision?
2. By what date must the decision be made?
3. Who has final authority?

Put those three answers on one page, share it, and the next “one more analysis” request has to meet a visible standard. That is not a rebuke. It gives the person asking a way to be diligent and decisive at once, which is what they wanted all along.

A short template:

| Question | Your answer |
| --- | --- |
| The decision we are making |  |
| What would change it |  |
| What we do if the data says A |  |
| What we do if the data says B |  |
| Deadline |  |
| Final decision owner |  |
| Default if no one decides |  |

If you cannot fill in the “A” and “B” rows, stop collecting and start discussing. The gap is not in the data.

## Making it safe to decide

A template will not fix the fear underneath. If the cost of being visibly wrong is high, people will keep reaching for more. Leaders can change that with a few small moves, none of which require a reorganization.

- **Say the pattern out loud, without a target.** “We’ve asked for more analysis on this three times. What are we protecting?” works better than “why are we stalling?”
- **Reward the stated decision rule, not just the outcome.** If a call was made with the evidence available at the time, praise the reasoning even when the result disappoints.
- **Make reversibility explicit.** “We can undo this in two sprints” takes pressure off the person who has to say yes.
- **Share ownership.** When a decision has one name on it, delay is the personal safe option. When the group owns the rule, committing is less lonely.

These are social moves, not analytical ones. They work because the thing holding the decision back was never a shortage of numbers.

## Before you build that course

The pattern shows up in an especially costly form in enablement. Performance dips, a target is missed, and two reflexes fight for the room. One says: we need more data on what’s going wrong. The other says: let’s build training. Both can be ways of avoiding the harder conversation about what is actually causing the gap.

The diagnosis step is where evidence earns its keep. The U.S. Department of Energy’s Lawrence Berkeley lab describes its [needs analysis process](https://training.lbl.gov/Resources/docs/NeedsAnalysisProcess.pdf) in four phases: identify performance expectations, determine whether a gap exists, identify the factors causing it, and decide the most effective and sustainable way to close it, whether through training, other changes or a combination. Training is one possible answer at the end, not the starting assumption.

![Four phases from expectations to performance response](https://media.babylovegrowth.ai/blog-images/organization-48833/1794548858987_Four-phases-from-expectations-to-performance-response.jpeg)

An open textbook on strategic HR management makes the same case from the manager’s side. A performance gap has to be connected to established responsibilities, behaviors, results or standards, not to a manager’s impression. And the response should follow the diagnosis: teaching, workflow redesign, extra resources, coaching or a formal process may each be right. It also observes that a knowledge test may confirm understanding without showing reliable performance under real working conditions ([Managing Poor Performance and Supporting Improvement](https://ecampusontario.pressbooks.pub/strategichrmgt/chapter/8-6-managing-poor-performance-and-supporting-improvement/)).

That is the line between useful and useless data in capability work. Useful evidence answers: is this a gap in what people can do, or a gap in the environment around them, such as unclear expectations, broken tools, conflicting incentives or a process that makes the right move hard? Another survey of learner satisfaction answers neither. Our guide to [skill gap analysis for HR leaders](https://cognistry.co/edge/skill-gap-analysis) walks through the diagnosis-first sequence, and the piece on [diagnosing skill, competency and capability](https://cognistry.co/edge/skill-vs-competency-vs-capability) helps when the terms themselves are blurring the conversation.

### What the evidence should come from

In an enterprise, the strongest grounding is the organization’s own source material: strategy documents, frontline friction, quality findings and the expertise of the people who do the work. Those sources tell you what the work actually requires. Three-month-old survey scores on a course nobody has built yet do not.

A good test for any evidence request in this space: if it comes back showing a capability gap, does the team know what it would build? If it comes back showing an environment problem, does the team know who would fix it? If neither branch has an owner, you are back at the Tangle.

### Use skills data with care

When the data is about people, the stakes for trust rise. MIT Sloan’s report on [how companies use AI to find and close skills gaps](https://mitsloan.mit.edu/ideas-made-to-matter/how-companies-can-use-ai-to-find-and-close-skills-gaps) describes how J&J’s leadership made clear that skills insights would not factor into performance reviews, that they would be de-identified and used in aggregate, and that employees could opt out. MIT CISR’s study of the same program adds that the company used skills inference only for personalized career development and aggregate strategic workforce planning, with other uses not permitted by its privacy function ([Resolving Workforce Skills Gaps with AI-Powered Insights](https://cisr.mit.edu/publication/2024_0401_DigitalTalentTransformation_VanderMeulenTonaLeidner)).

The lesson for leaders is about permission and purpose: say what the data is for before you gather it, and people are far less likely to treat the request as a threat. The question can also be very specific to a trade. In technical settings, where skills such as controller programming and troubleshooting are built by practice, [industrial automation training](https://onlineplcsupport.com/industrial-automation-training/1793366463941_isometric-platforms-suggesting-practical-automation-learning) shows how narrowly a capability can be defined, which makes the question “what would we do with the answer?” much easier to ask.

## A short field guide for the next meeting

When the request comes, try this sequence. It takes about five minutes.

1. **Name the decision.** “We are deciding whether to redesign the onboarding flow or leave it.”
2. **Ask the two-sided test.** What would we do if it came back high? What if low?
3. **Check reversibility.** If we are wrong, how fast can we change course?
4. **Set the box.** A date, an owner, and the single question the data must answer.
5. **Agree the default.** If nothing changes by the date, what happens?

If the request survives all five, it is a good request. Collect the data, and hold everyone, including yourself, to the date.

> We have watched a lot of good teams wait on the next analysis. Rarely was the evidence thin. Usually it was that no one wanted to be the first to say, “this is enough.”

## Promo

If your organization is deciding whether a performance problem calls for training at all, the Cognistry Platform starts earlier than most tools. It diagnoses what capability the work actually requires, decides whether learning is the right response, and grounds that decision in your own evidence: strategy, frontline friction, quality findings and subject-matter expertise. From there it structures the response and measures outcomes back to the business. See how it works on the [Cognistry Platform](https://cognistry.co/platform) page.

![Cognistry](https://csuxjmfbwmkxiegfpljm.supabase.co/storage/v1/object/public/blog-images/organization-48833/1785714030894_cognistry.jpg)

## FAQ

### How do I know if “I need more data” is a stall or a real need?

Ask what decision the data would change and what you would do if it came back either way. If the person can name both branches, the request is real. If they cannot, the data is unlikely to move the decision, and the conversation should shift to what is making the call feel risky.

### How much data is enough to decide?

Decide in advance. The National Academies guide suggests defining enough as, for example, answering one key question or covering 75% of a list. Scale it to the stakes: reversible decisions need less evidence than ones you cannot undo.

### Is asking for more data a sign of weak leadership?

No. It is a sane response in systems that punish visible mistakes, and it often comes from people who care about getting it right. The task for a leader is to change what is safe: make reversibility explicit, share ownership of the decision and credit sound reasoning.

### How does this apply to training and enablement decisions?

Before building a course, confirm that the gap is one learning can close. That means grounding the question in organizational evidence about what the work requires and where it breaks down, then deciding on the response. Our guide to [six-step competency mapping](https://cognistry.co/edge/competency-mapping-guide) shows one way to structure that diagnosis.

## Sources

- [How to Make Better Decisions with Less Data, The Conference Board](https://www.conference-board.org/research/human-capital-briefs/blg-00-1-5304)
- [Advancing the Art and Science of Decision-Making, National Academies](https://www.nationalacademies.org/read/29265/chapter/13)
- [Deciding to Decide: How Enterprise Organizations Mistake Data Collection for Progress, MporgSoft](https://mporgsoft.com/deciding-to-decide-enterprise-data-collection-analysis-paralysis/)
- [Data-Driven Decision Making Still Needs a Method, Sufficient Certainty](https://www.sufficientcertainty.com/data-driven-decision-making)
- [Today’s workforce challenges demand data diversity and diligence, Deloitte](https://action.deloitte.com/insight/ucyfegbtzqfrzdh4rnmy6dhn/todays-workforce-challenges-demand-data-diversity-diligence)
- [Needs Analysis Process, Lawrence Berkeley National Laboratory](https://training.lbl.gov/Resources/docs/NeedsAnalysisProcess.pdf)
- [Managing Poor Performance and Supporting Improvement, eCampusOntario](https://ecampusontario.pressbooks.pub/strategichrmgt/chapter/8-6-managing-poor-performance-and-supporting-improvement/)
- [How companies can use AI to find and close skills gaps, MIT Sloan](https://mitsloan.mit.edu/ideas-made-to-matter/how-companies-can-use-ai-to-find-and-close-skills-gaps)
- [Resolving Workforce Skills Gaps with AI-Powered Insights, MIT CISR](https://cisr.mit.edu/publication/2024_0401_DigitalTalentTransformation_VanderMeulenTonaLeidner)
- [Industrial automation training, Online PLC Support](https://onlineplcsupport.com/industrial-automation-training/1793366463941_isometric-platforms-suggesting-practical-automation-learning)

## Recommended

- [The Hidden Cost of Data Drag](https://cognistry.co/edge/the-hidden-cost-of-data-drag)
- [COO Decision Systems](https://cognistry.co/edge/coo-decision-performance-system)
- [What Higher-Value Work Costs When Nobody Defines It](https://cognistry.co/edge/what-higher-value-work-costs-when-nobody-defines-it)

[Prove Value](https://cognistry.co/edge/tag/prove-value)

## Continue Reading

[![Skills Ontology for HR Leaders: A Practical Guide](https://csuxjmfbwmkxiegfpljm.supabase.co/storage/v1/object/public/blog-images/organization-48833/1785803661731_Decorative-title-card-illustration-for-skills-ontology-article.jpeg)](https://cognistry.co/edge/skills-ontology)

Prove Value

### [Skills Ontology for HR Leaders: A Practical Guide](https://cognistry.co/edge/skills-ontology)

 Skills Ontology for HR Leaders: A Practical Guide A skills ontology is a structured knowledge...

Brian Lambert, PhD · 28 min read

[![Future Skills: Start With Contributions, Not Trait Lists](https://media.babylovegrowth.ai/blog-images/organization-48833/1792186247552_Hand-drawn-workflow-objects-frame-the-title.jpeg)](https://cognistry.co/edge/future-skills)

Develop People

### [Future Skills: Start With Contributions, Not Trait Lists](https://cognistry.co/edge/future-skills)

 Show me how you’d assess “critical thinking” on Monday. Not define it. Assess it, in a real...

Mark Ondash CPTD® MPC™ · 21 min read

[![](https://media.babylovegrowth.ai/blog-images/organization-48833/1793086899882_Sketch-motifs-framing-the-article-title.jpeg)](https://cognistry.co/edge/the-difference-between-capabiliity-engineering-and-hr)

Train Agents

### [Capability Engineering vs. HR: What Sets Them Apart](https://cognistry.co/edge/the-difference-between-capabiliity-engineering-and-hr)

 HR manages the people system: hiring, pay, performance processes, employee relations and talent...

Brian Lambert, PhD · 24 min read

[![Cognistry\_Icon\_White](https://cognistry.co/hs-fs/hubfs/Logos/Cognistry_Icon_White.png?width=150&height=60&name=Cognistry_Icon_White.png "Cognistry_Icon_White")](https://cognistry.co)

- [Cognistry](https://cognistry.co)
- [Resources](https://cognistry.co/resources)
- [Sign In](https://app.cognistry.co)
- [Privacy](https://cognistry.co/privacy)
- [Terms](https://cognistry.co/terms)

© 2026 Cognistry. Capability by Design.

```json
{
  "@context" : "https://schema.org",
  "@type" : "BreadcrumbList",
  "itemListElement" : [ {
    "@type" : "ListItem",
    "item" : "https://cognistry.co/edge",
    "name" : "Insights",
    "position" : 1
  }, {
    "@type" : "ListItem",
    "item" : "https://cognistry.co/edge/tag/prove-value",
    "name" : "Prove Value",
    "position" : 2
  }, {
    "@type" : "ListItem",
    "name" : "When “I Need More Data” Becomes a Safe Place to Hide",
    "position" : 3
  } ]
}
```

```json
{
  "@context" : "https://schema.org",
  "@type" : "BlogPosting",
  "author" : {
    "@type" : "Person",
    "name" : "Cognistry Team",
    "url" : "https://cognistry.co/edge/author/cognistry-team"
  },
  "dateModified" : "2026-11-18T17:32:30.864Z",
  "datePublished" : "2026-11-15T10:45:00.000Z",
  "headline" : "When “I Need More Data” Becomes a Safe Place to Hide",
  "image" : [ "https://media.babylovegrowth.ai/blog-images/organization-48833/1794548747753_Hand-drawn-evidence-and-decision-making-title-card.jpeg" ],
  "mainEntityOfPage" : {
    "@id" : "https://cognistry.co/edge/i-need-more-data",
    "@type" : "WebPage"
  },
  "publisher" : {
    "@type" : "Organization",
    "logo" : {
      "@type" : "ImageObject",
      "url" : "https://cognistry.co/hubfs/Logos/Cognistry%20Logo.png"
    },
    "name" : "Cognistry Digital Command"
  }
}
```