Redefining trust in data accuracy

Ensuring data quality is more than a compliance requirement—it is the foundation of effective decision-making, accountability, and program integrity. In the context of large-scale health programs and donor-funded initiatives, unreliable data can lead to misguided policies, inefficient resource allocation, and inaccurate impact assessments. This document provides a comprehensive, structured framework for conducting Data Quality Audits (DQA), equipping M&E professionals, program managers, and auditors with the methodologies necessary to verify data accuracy, assess reporting systems, and strengthen data management practices. Built around a step-by-step auditing methodology, this guide details the full process of conducting a DQA, from selecting indicators and auditing sites to verifying reported data and evaluating data management systems. It introduces key dimensions of data quality, including accuracy, completeness, reliability, timeliness, integrity, and confidentiality, ensuring that reported results align with reality. The document provides clear protocols for tracing data from service delivery points to national reporting systems, identifying discrepancies, inconsistencies, and gaps that may compromise program credibility. For M&E specialists, policymakers, and implementing partners, this resource is an essential tool for ensuring that reported program results are not just numbers but accurate reflections of impact. By following the structured processes outlined in this guide, organizations can strengthen data verification mechanisms, improve reporting consistency, and enhance accountability to donors and stakeholders. A robust data quality auditing system is not just about validation—it is about building trust, improving program effectiveness, and ensuring that every decision is informed by reliable evidence .

📖 Let me tell you a story of how I think we can solve the data trust and quality crisis we face today... 📖 Imagine this: Your company has just launched a new data product. Everyone is excited, the KPIs look great, and users are relying on it for key business decisions. But soon, questions start popping up. "Why don’t these numbers match what we saw last quarter?" "Are these KPIs based on solid data?" The data team assures them that the numbers are correct—but they know the reality. Behind the scenes, data quality isn’t always perfect, and sometimes they’re forced to deliver results based on optimistic estimates. The trust gap begins to grow. This is where the Trust-Tiered Interfaces pattern comes into play. 💡 With this approach, instead of delivering one opaque interface, the product offers users three clear choices: - High Confidence Interface 🔒: Where users get only the rock-solid, validated data—perfect for making high-stakes decisions with confidence. - Optimistic Interface 🌟: Optional, but more comprehensive, where corrected data is included. It gives a broader view, while still based on accurate info. - Data Quality Interface 🔍: Here's the game-changer—an interface that shows exactly how reliable the data is. It’s fully transparent about the sources, gaps, and uncertainties, so users know what they’re dealing with. Before this, most teams offered either the high confidence or optimistic view without giving users insight into data quality. But hiding those imperfections was a loophole—one that quietly allowed issues to slip from one data product to another. 🔑 Here’s the truth: Data will never be perfect, and that’s okay! The key is being upfront about it. By offering the Trust-Tiered Interfaces, data teams can empower users to understand the quality of the data they’re working with. This increases trust not only in the data but in the product and the team itself. Imagine a world where every business decision is made on the right data, with full awareness of its limitations. That’s the kind of maturity this pattern can bring. #DataProducts #DataMesh #DataManagement

One of the most popular false statements I hear is "issues with trust in data are mostly technical.” Examining the Statement - Common Belief: Trust problems vanish if we fix errors in code, tools, or pipelines. - Key Question: Can flawless data still be mistrusted if people don’t understand or believe in its source or purpose? Rethinking “Trust” Trust isn’t just about accuracy; it’s about transparency, context, and credibility. Even perfect data won’t be trusted if no one knows where it came from or why it matters. Think of a beautifully wrapped gift from a stranger. Without knowing what’s inside or who sent it, scepticism persists. In Practice - Documentation & Explanation: Show how data is collected, validated, and maintained. - Open Communication: Invite questions and be honest about limitations. - Cultural Acceptance: Foster an environment where challenges to data are welcomed and addressed. Trust in data isn’t earned by technical perfection alone. It grows when people understand, relate to, and find meaning in the numbers they rely on. #DataTrust

Is “good enough” data really good enough? For 88% of MOps pros, the answer is a resounding no. Why? Because data hygiene is more than just a technical checkbox. It’s a trust issue. When your data is stale or inconsistent, it doesn’t just hurt campaigns; it erodes confidence across the org. Sales stops trusting leads. Marketing stops trusting segmentation. Leadership stops trusting analytics. And once trust is gone, so is the ability to make bold, data-driven decisions. Research tells that data quality is the #1 challenge holding teams back from prioritizing the initiatives that actually move the needle. Think of it like a junk drawer: If you can’t find what you need (or worse, if what you find is wrong), you don’t just waste time, you stop looking altogether. So what do high-performing teams do differently? → They schedule routine maintenance. → They establish ownership - someone is accountable for data processes. → They invest in validation tools - automation reduces the manual grind. → They set governance policies - because clean data only stays clean if everyone protects it. Build a culture where everyone values accuracy, not just the Ops team. Because clean data leads to clearer decisions and a business that can finally operate with confidence.

It’s rewarding to see a nearly 200-year-old institution reimagine itself for the AI era. During a recent conversation with the team at Citizens Bank, I was impressed by how they're shifting data management from a back-office function into a strategic competitive advantage. This example also reminded me that the most powerful transformations happen when you build on a foundation of trust.     Citizens took a bold approach with their master data management (MDM) modernization. They moved from batch processing that took days to near real-time data synchronization across their 1,000+ branches in 14 states. Using Informatica's Intelligent Data Management Cloud (IDMC) platform on Amazon Web Services (AWS), they've reduced data onboarding time by approximately 85% and transformed MDM into what they call a "Tier 1" operational asset, meaning it’s always available, accurate and ready to power customer interactions.    The results speak volumes. What used to take three days — even something as simple as updating a customer's phone number — now happens instantly. Their contact center call volumes decreased, their mobile experience became seamless and every customer interaction now draws from a single, trusted source of truth.    What I find particularly compelling is how Anand Vijai M R and his team built flexibility into their architecture while maintaining consistency across every customer touchpoint. The cloud-native approach freed their teams from infrastructure complexities so they could focus on what truly matters: ensuring data accuracy and powering AI use cases. With CLAIRE as an AI copilot, they've democratized access to trusted data across the organization.    This is the transformation I'm seeing across industries. Organizations that treat data as a strategic platform are building sustainable competitive advantages in the AI era. For a bank with roots dating back to 1828, Citizens proves that innovation and tradition can coexist harmoniously. https://lnkd.in/gZCua-M9  

“If nobody trusts your #CMDB, it doesn't matter how accurate it is.” Imagine having the world’s most advanced medical scanner… It’s fast, precise, 99% accurate but if doctors don’t trust the readings, they’ll never use it to make life-saving decisions. That’s what a clean but untrusted CMDB feels like in #IT operations. You can have: - 90% data accuracy - All the CI relationships mapped - Owners assigned - Discovery tools humming... But if: ⚠️ No one references it in Change, Incident and Problem ⚠️ DevOps ignores it ⚠️ Support teams question the ownership …it becomes shelfware with a heartbeat... impressive, but irrelevant. Data quality ≠ Data usability. Data isn’t valuable until someone relies on it. So how do you build trust in the CMDB? Not with tools but with culture: - Visibility: Make it part of workflows, not a silo. - Stewardship: Assign owners who own and evolve data. - Accountability: Align SLAs to CI health, not just ticket closure. - Application: Use it in Change risk scoring, Incident impact, and AI Ops correlation. Even the most intelligent #AI can’t operate on data your people don’t believe in. AI isn’t just about ingesting data. It’s about acting on trusted data. So here’s the question. Who owns CMDB trust in your organization? Is it the tool or the people behind it? #CMDB #CSDM #AIOps #ServiceNow #DigitalTransformation #Data #ITOperations #Leadership #ITSM #Strategy

What "I don't trust the data" actually means: - "The data doesn't match my expectations" - "The data contradicts my experience" - "I don't understand how the data was collected" - "I've been burned by incorrect data before" - "I don't know the limitations of this analysis" - "I have information the data doesn't capture" - "The data threatens my position or authority" - "I'm not ready to change based on what I'm seeing" Data trust isn't just about accuracy. It's about: - Psychological safety - Transparent processes - Consistent definitions - Acknowledged limitations - Aligned incentives - Respected expertise Data quality matters, but even perfect data will be rejected if these human factors aren't addressed. Building trust requires more than validation. It requires vulnerability, empathy, and patience.

A big part of building Motif Analytics has been re-thinking how an exploration-first analytics tool should work. Most analytics tools today are built for narrow reporting, provide appealing but misleading insights, or require many hours of work by a strong data practitioner to answer practical custom questions. We asked ourselves: how can we adjust existing analytics abstractions and tradeoffs to provide the fastest path from raw data to practical insights in modern tech companies? This is hard because many fundamental assumptions about how analytics “ought to be done” have become entrenched and are rarely questioned. I’d like to do a series of short posts about several such assumptions, which can get in the way of fast practical analytics, and hear reactions from data folks (you!). Let’s start with one of the biggest headaches of every data practitioner I know. 🔬 Data Quality 🔬 Organizational trust in the accuracy of metrics is something every data practitioner has to grapple with. The widely accepted solution is improving and monitoring data quality through every step of data capture and processing. But does it solve the problem? Do you know 100+ person organizations where trust in metrics is not an issue? Several big factors work against it: ➡ no guarantee of result correctness: missing even one small thing breaks the whole processing chain ➡ dynamic environment: software products and logging are constantly changing, business definitions of metrics are shifting ➡ inherent errors: there is an inherent loss of analytics logs (often ~1%) ➡ distributed ownership: feature engineers, data engineers, analysts and data scientists all touch the same data Is there another approach? How do strong data folks answer analytics questions today when working with imperfect data? They: 1️⃣ check results correctness through their coherence over: - time: review metric stability over time - context: view data in broad context of prior/later behavior - question tweaks: inspect how results change based on slight changes to the question - redundancy: compare metric values coming from redundant data sources. 2️⃣ work around identified data quality issues quickly during the analysis by filtering out bad data, using proxy data, making reasonable simplifying assumptions, etc. Both are specific to the question at hand and hard to generalize across analyses. Unfortunately, analytics tools today don't focus on making this type of work easy. Here are some approaches we are using for Motif: ➡ display data in broad context including prior/later events in user flows ➡ preserve high interactivity with ~2 second exploratory query time on any dataset sizes ➡ provide the ability to modify data on the fly by replacing event patterns.

Recent client meetings have left me a bit stumped! Because I keep hearing the following: “We don’t trust our data.” It's not the first time I've heard it, and I bet it won’t be the last. The irony? Those same businesses were using data every single day to pay invoices, run supply chains, and make strategic calls. So it’s not really the data they mistrusted. It must be something deeper. So where does this mistrust come from? Sometimes it’s a cover for not liking what the numbers say (because numbers don’t bend to opinion). Other times, it’s really about trust in the data team rather than the data itself. Occasionally, it’s just become a lazy throwaway line. If organisations want to break this cycle, both leaders and data teams need to change the way they work together. Here’s a 5 point playbook that stops “data mistrust” in its tracks: 1. Define Once, Use Everywhere: agree common definitions for key metrics. Document them, make them visible, and hold teams accountable for sticking to them. Consistency builds confidence. 2. Show the Journey: make data lineage transparent. Leaders should see where a number originates, how it’s transformed, and why it ends up in a dashboard etc. Traceability removes suspicion. 3. Shared Accountability: data isn’t an “IT product.” It’s a joint effort. Business leaders must own the accuracy of inputs; data teams must own the quality of models and outputs. Co-ownership prevents finger-pointing. 4. Resolve Issues Quickly: don’t let data concerns fester. Implement visible feedback channels, track issues openly, and close them with clear communication. The faster issues are addressed, the stronger trust becomes. 5. Normalise Hard Truths: not all insights will be comfortable. That’s the point. Leaders must be ready to hear what the numbers say, and data teams must present them clearly. Data itself isn’t untrustworthy. It’s the behaviours, mindset, and responses around it that determine whether people believe it. So let’s stop hiding behind the lazy phrase “we don’t trust our data.” 👉 Business leaders are you really questioning the data, or just avoiding what it’s telling you? 👉 Data teams are you giving the business clarity, speed, and confidence, or just more numbers to argue over? Because until both sides stop passing the blame, “data mistrust” won’t go away, it will just keep undermining decisions. Mark Stouse Bill Schmarzo Malcolm Hawker Eddie Short Kyle Winterbottom Edosa Odaro Joe Reis Matthew Small Dan Everett

In the realm of artificial intelligence, discerning truth from falsehood is more than a philosophical question—it’s a practical challenge that impacts business decisions and consumer trust daily. We are designing our new systems inspired by the classic dilemma of the Village of Truth and Lies, that can reliably manage the accuracy of their outputs. Here are some practical approaches that we are finding useful. 1. Multiple Agents: Use different AI models to answer the same question to cross-verify responses. 2. Consistency Checks: Follow-up with related questions to check the consistency of AI responses. 3. Confidence Estimation: Measure how confident an AI is in its answers, using this as a heuristic for reliability. 4. External Validation: Integrate verified databases to confirm AI responses wherever possible. 5. Feedback Loops: Incorporate user feedback to refine AI accuracy over time. 6. Adversarial Testing: Regularly challenge the system with tough scenarios to strengthen its discernment. 7. Ethical Responses: Design AIs to admit uncertainty and avoid making up answers. 8. Audit Trails: Keep logs for accountability and continuous improvement. I am also looking at game theoretic approach to estimating AI confidence. If you are interested in learning more, please feel free to connect for a discussion. Managing accuracy and trust is critical factor. By crafting smarter, self-aware AI systems, we pave the way for more reliable, transparent interactions—essential in today’s data-driven landscape. Please share your thoughts in the comments. #ArtificialIntelligence #MachineLearning #DataIntegrity #BusinessEthics #Innovation