Improving Data Foundation for Nonprofit Analytics

Data and analytics in the nonprofit sector in the UK is honestly wanting, so I decided to do something about it. I've been working on something for the past several months; a practical framework/guide to use when conducting data and analytics audits in the nonprofit and charity sector in the UK and beyond. It came from a pattern I've been seeing across organisations: large amounts of data being collected, reports being produced, and very little clarity about whether the numbers were accurate, what they actually meant, or whether the people the organisation was serving had any voice in them at all. I have almost a decade of data and M&E experience across NGOs, government, and the private sector in Kenya and the UK. I've noticed that the problem is rarely a lack of data. It is almost always a lack of data infrastructure, definitions, governance, and honest measurement. The framework I've been working on covers audit methodology, data quality assessment, outcomes measurement, maturity modelling, and implementation planning. It will be free for anyone working in the nonprofit sector to download and use as a guide when conducting data audits. Whether you're a data analyst, a CEO, or a programme manager who inherited a reporting mess, this framework is for you. I'll be publishing it soon. If this is something your organisation needs, or you know someone who does, share this post. Let's get it to the people who need it most. And follow to be notified when I publish. Also, do you work in a nonprofit or public sector, how is the data ecosystem? I know some nonprofits don't even have data functions which is alarming. #dataanalysis #nonprofit #datagovernance #charitysector #dataanalytics #monitoringandevalution

𝗙𝗼𝗿 𝗺𝗮𝗻𝘆 𝗻𝗼𝗻𝗽𝗿𝗼𝗳𝗶𝘁𝘀, "𝗦𝗮𝗹𝗲𝘀𝗳𝗼𝗿𝗰𝗲" 𝗶𝘀𝗻'𝘁 𝗮 𝗖𝗥𝗠. It's a collection of 5,000+ duplicate contacts, 7 fragile integrations, and 4 critical reports that are just exported to a spreadsheet anyway. The vision is a "single source of truth." The reality is manual receipts and data entry. We all know the friction points:  • 𝗗𝗮𝘁𝗮 𝗙𝗿𝗶𝗰𝘁𝗶𝗼𝗻: Duplicates from every import. Addresses that age faster than you can clean them.  • 𝗣𝗿𝗼𝗰𝗲𝘀𝘀 𝗙𝗿𝗶𝗰𝘁𝗶𝗼𝗻: Using spreadsheets as a database. Using your inbox as a coordination tool.  • 𝗧𝗲𝗰𝗵 𝗙𝗿𝗶𝗰𝘁𝗶𝗼𝗻: That online donation integration that fails silently, losing data until a fundraiser complains. 𝗛𝗲𝗿𝗲'𝘀 𝘁𝗵𝗲 𝗯𝗿𝘂𝘁𝗮𝗹 𝘁𝗲𝗻𝘀𝗶𝗼𝗻: They are selling you a "unified data model." But you can't unify chaos. You can't "orchestrate" with Flow and Omni-Channel when you don't even know what data is trustworthy. But here's what actually works, with measurable outcomes: 𝗙𝗜𝗫 #𝟭: 𝗗𝗮𝘁𝗮 𝗦𝘁𝗮𝗯𝗶𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻  → Implement rule-based matching and duplicate jobs  → 80% reduction in duplicates (50k database: 4,000 down to 800)  → 8% increase in campaign conversion rates  → Measurable fundraising ROI improvement 𝗙𝗜𝗫 #𝟮: 𝗙𝗹𝗼𝘄-𝗕𝗮𝘀𝗲𝗱 𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗼𝗻  → Automate receipts, reconciliation, exception reporting  → 60% reduction in manual workload  → 50% fewer errors from manual processes  → SLAs drop below 24 hours 𝗙𝗜𝗫 #𝟯: 𝗦𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲𝗱 𝗢𝘂𝘁𝗰𝗼𝗺𝗲 𝗧𝗿𝗮𝗰𝗸𝗶𝗻𝗴  → Build results model with 3-5 core indicators  → 70% reduction in report creation time  → Real-time KPI visibility for leadership  → 40% improvement in volunteer coordination efficiency Forget the 5-year AI roadmap. The real, un-glamorous work is 𝘀𝘁𝗮𝗯𝗶𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻. Before you look at another new feature, go find the one process that breaks the most. See why it breaks. Fix just that. Discipline before orchestration. Visibility before migration. That's the real roadmap. I'd love to hear from the Admins and Consultants working on Nonprofit Cloud. Share your thoughts on this (Maybe I'll learn something I am missing). #Salesforce #SalesforceAdmin #Nonprofit #NPSP #NonprofitCloud 

The latest episode of the Fundraising.ai podcast was recorded live at AWS HQ in New York, and it felt like one of those conversations that cuts through the noise fast. For the past few years, nonprofits have been nudged to believe AI is the strategy. Adopt the right tools quickly enough and the future will take care of itself. But that framing is backwards - or at least incomplete. In this conversation, industry pioneer and dear friend Kelley Hecht (AWS) and I unpack a more grounded truth: AI is not the strategy. It is a powerful set of tools that will either amplify clarity or accelerate chaos. Here are 6 takeaways we dive into that, I believe, can make a big difference for nonprofits navigating their AI journey: 1️⃣ Start with outcomes, then work backwards (mission obsessed, constituent obsessed). 2️⃣ Data strategy is not optional anymore - it is the foundation AI stands on. 3️⃣ Perfect data is a myth. Progress beats waiting to be “ready.” 4️⃣ Analytics should not only be a rearview mirror - it should help you drive. 5️⃣ This is a team sport. Real traction takes cross-functional ownership. 6️⃣ Senior leadership support matters (and CFOs can be an underrated ally). If you have felt overwhelmed by AI conversations, unsure where to start, or worried you are falling behind without even knowing what “ahead” looks like, this episode is for you. Listen here: https://lnkd.in/ef4BmFae Kelley brings what she always brings: intention, wisdom, and strategy (plus just enough friendly skepticism to keep the hype in check). Also, she now holds the unofficial record for most Fundraising.AI return visits, so yes, we probably owe her a jacket. #FundraisingAI #Nonprofits #DataStrategy #ResponsibleAI #AWS #CloudComputing #NonprofitLeadership #AIForGood

How do not-for-profit organizations overcome barriers to scale on tight budgets? Andrew Patricio, Principal for Data and Analytics at UnidosUS (@WeAreUnidosUS) shared their story and the answer to today’s big question. The story starts here: “I walked into a sea of Excel—nothing but spreadsheets.” UnidosUS is the largest Latino civil rights and advocacy organization in the U.S., supporting more than 300 affiliate nonprofits in health, education, housing, and workforce development. Their mission is expansive, but their tools weren’t keeping pace. Affiliate data was locked in a 256-column spreadsheet. One team fielded every data request. And every answer required time someone didn’t have. The risk wasn’t just inefficiency. It was a missed opportunity to act faster, listen closer, and deliver greater impact. That’s why Andrew sought out Quickbase.  “It’s been my go-to tool for years. Especially when you’ve got a pile of spreadsheets and no time, budget, or need for a full custom app—Quickbase is perfect.” Andrew rebuilt their affiliate system in Quickbase. Data became accessible without compromising control. Reporting became self-service. And for the first time, the organization could see and scale what was working, without needing a full-time developer. The impact speaks for itself: $80K+ saved in software spend, real-time visibility across affiliates, a more confident, empowered team, and a new way of thinking about what tech should do. UnidosUS doesn’t exist to be good at software. Chances are, neither does your organization. → When your productivity tools let you work the way you want to work, you get to spend more time scaling the impact of your mission and less time fighting your tech stack.  https://lnkd.in/d_ksmmfp 

What an “AI-Ready Organization” Actually Looks Like (And why most institutions aren’t there yet.) Over the last 5 years, I’ve worked with nonprofits, universities, and mid-size organizations that want AI but struggle to get real outcomes from it. The reason is simple: AI doesn’t fail because of the model. It fails because the organization isn’t ready for it. Here’s what readiness actually looks like (with the numbers behind it): 1. Clean, trustworthy data Only 12% of organizations say their data is AI-ready and 67% don’t fully trust the data they use for decisions. AI can’t compensate for inconsistent, siloed, or outdated information. 2. Documented, stable workflows Where workflows are mapped and standardized, organizations see 30–50% cost reduction and major drops in error rates when they apply AI. AI accelerates clarity or accelerates chaos. 3. Internal champions, not just tools 77% of employees already have the potential to become AI champions. The companies that scale AI make these people the engine of adoption. Capability beats top-down mandates. 4. Governance that enables scale 62% of organizations say the biggest barrier to AI is lack of governance, not lack of technology. Clear roles, data access rules, and risk guidelines make AI reliable and safe. If your organization wants to see real results from AI, the foundations matter: your data has to be reliable, your workflows have to be clearly defined, your teams need the capability to adopt new ways of working, and your governance must support responsible scale. These elements together create the operational maturity that turns AI from a series of experiments into measurable outcomes. If you’re building this maturity within your institution, whether a nonprofit, a university, or an enterprise, I’m happy to share what I’m seeing across the sector and what’s actually working. Book a free consultation call. Link in comments.

Unlock the power of social impact measurement with the Actionable Impact Measurement (AIM) framework. It's a step-by-step guide, a compass for your mission-driven organization, from Groundwork to Impact Reporting. First, Groundwork. We'll facilitate the essential discussions and documentations needed to set up your impact framework. It's about laying a solid foundation, aligning your Vision, Mission, and Goals with your Program Structure and Theory of Change. Next, Metrics. They're the lifeblood of your impact goals. The AIM guidebook will help your organization select meaningful and practical metrics to bridge the gap between funders, intermediaries, and grantees/investees. Then, Data. It's the backbone of your impact strategy. Unearth the potential of technology in collecting reliable and credible data on your selected outcome-oriented metrics. Finally, Impact Communication. Tying it all together is your organization's impact story. Learn methods to communicate it effectively, using the right metrics, transparent data, and minimal tools. But that's not all. You'll also delve into social impact indicators, data capacity improvement, and become familiar with terms like Theory of Change, Logic Model, LogFrame, and Impact Thesis. The AIM guide has already helped many over 5000+ organizations. Now, it's your turn. Download the free AIM guide. Start your journey towards effective social impact measurement today. Because the world needs more organizations that not only aim to make a difference but know how to measure it. ------------------------------------------------------------------------------- If this strikes a chord with you, FOLLOW Unmesh Sheth Sopact Every weekday, I'm unwrapping insights to turbocharge your data-driven nonprofit journey. Let's collaborate on creating data-driven culture for mission driven organizations. Dive into Discussion: Imagine the possibilities this could unlock for us. 💬 Comment 🔁 Repost ❤️ Like Let's make an impact, together. #socialimpact #nonprofit #impactmeasurement #monitoringandevaluation #socialenterprise #donation #fundraising

Putting pressure on data science teams to deliver analytical value with LLMs is cruel and unusual punishment without a scalable data foundation. Over time, the best LLMs will be able to write queries as effectively or more effectively than an analyst - or at minimum make writing the query easier. However, the most cost-intensive aspect of answering business questions is not producing SQL, but deciding what the query inputs should be and determining whether or not the inputs are trustworthy. Thanks to the rapid evolution of microservices and data lakes, data teams find themselves living in a world of fragmented truth. The same data points might be collected by multiple services, defined in multiple different ways, and could actually be going in opposite and contradictory directions. Today, data developers must do the hard work of understanding and resolving those discrepancies, which comes in the form of 1-to-1 conversations with the engineers managing logs and databases. Very few if any service teams at a company have documented their data for the purpose of analytics. That results in a giant gap in documentation across 1000s of datasets across the business. Without this gap being filled, data scientists will ultimately have to manually hand-check any prediction that an LLM makes in order to ensure it is accurate and not hallucinating. The model is doing a job with the information it has, but the business is not providing enough information for the model to deliver trustworthy outcomes! By investing in a scalable data foundation, this paradigm flips on its head. Data is well documented, clearly owned, and structured as an API enforced by contracts that define the use case, constraints, SLAs, and semantic meaning. A quality-driven infrastructure is a subset of all data in the lake, which reduces the surface area LLMs need to make decisions only to the nodes in the lineage graph which have clear governance and change management. Here's what I suggest: 1. Start by identifying which pipelines are most essential to answering the business's most common questions (you can do this by accessing query history) 2. Identify the core use cases (datasets/views) that are leveraged in these pipelines, and which intermediary tables are of critical importance 3. Define semantically what the data means at each level in the transformation. A good question to ask is "What does a single row in this table represent?" 4. Validate the semantic meaning with the table owners 5. Get the table owners to take ownership of the dataset asn API, ideally supported programmatically through a data contract 6. Define the semantic meaning and constraints within the data contract spec, mapped to a source file 6. Limit any usage of an LLM to the source files under contract Good luck! #dataengineering

What happens to data collection when funding feels uncertain? I am hearing this question a lot lately: “We just lost our federal funding, so we will pause all our data collection projects” or “What if the next round of cuts hits us? So we have halted all new data projects.” All understandable…yes. Fear around funding often makes data collection the first thing to pause. But ironically, data is also what makes your case stronger, my nonprofit friends—with communities, with funders, and with the people you are trying to reach. The goal with your data collection projects isn’t to collect more data; it is to collect better data. So, what can you do when you have funding threats and still need to continue your data collection work? Think of it as a recipe for data listening: A. If you haven’t collected data recently: ● Start small. A short survey, a focus group, or even a few intentional conversations with community members. ● Ask open questions like “What matters most to you right now?” rather than a checklist of satisfaction ratings. B. If you just ran a survey in the last 6-10 months: ● Don’t rush to collect more. Instead, sit with what you already have. Re-read open-ended responses. ● Ask: Did we actually close the loop with our community and tell them what we learned? ● Share back 2–3 insights and invite new community reflection: “Does this reflect your reality?” That is listening, too. C. If your data feels disconnected from action: ● Pick one theme or finding. Use it to shape one program tweak or one donor conversation. ● Keep track of how that single change shifts engagement—proof that your data is working for you, not just sitting in a file. Funding will ebb and flow. But building trust through data is always worth it. And if you need support: look in the comments :) #nonprofits #nonprofitleadership #community #fundraising

No Data = No Value: The Inconvenient Truth About AI Everyone wants to talk about AI models. Nobody wants to talk about data. That’s a problem. The Uncomfortable Reality The most powerful algorithm in the world, fed garbage data, produces garbage outcomes. No exceptions. Yet organizations keep pouring millions into model development while treating data quality as an afterthought. Data scientists spend 60–80% of their time just cleaning data. That tells you everything you need to know about the state of most organizations’ data. What Trustworthy AI Actually Requires ·       Accurate data — or your model is confidently wrong ·       Complete data — or your model has dangerous blind spots ·       Consistent data — or your model learns confusion ·       Timely data — or your model solves yesterday’s problems ·       Properly labeled data — or your model systematically fails Miss any one of these, and trust evaporates. The Stakes Are Real Flawed hiring tools. Biased customer scoring. In every high-profile AI failure, the root cause wasn’t a bad algorithm. -->It was bad data. And with state-level AI legislation gaining momentum, “we didn’t know our data was flawed” won’t cut it anymore. What We’ve Seen at PwC This isn’t theory—it’s what we’ve lived. At PwC, our experience has been clear: investing in strong data governance, quality, and management didn’t just improve our AI outcomes—it skyrocketed the speed and quality of our AI delivery. When the data foundation is right, everything accelerates. Models train faster. Results are more reliable. Stakeholders trust the outputs. Deployment timelines shrink dramatically. When it’s not, even the best teams spend their time firefighting data issues instead of building value. What To Do About It 1.     Fund data like you fund models. Focused data investment in not optional. 2.     Treat labeling as strategy, not grunt work. Labelers encode judgment into your system. 3.     Measure data quality as a KPI. Not just model accuracy. 4.     Build feedback loops. Data drifts. Catch it early. The Bottom Line The winners in AI won’t have the fanciest models. They’ll have the cleanest data. Poor data = poor outcomes. Quality data = trustworthy AI. Stop chasing algorithms. Start fixing your data. Have you seen AI projects fail because of data quality? Share your experience below. #ArtificialIntelligence #DataQuality #TrustworthyAI #ResponsibleAI #AIStrategy Chris Dulny Jennifer Johnson Alaina Tennison, FCPA, FCA Josée St-Onge, FCPA, CIA, CRMA, PMP Mitchell Reisler Robert Bzdziuch Vanessa H.

A nonprofit recently told me: “We have the data. We just don’t know where to fund next.” They were running multiple programs across regions: education, livelihoods, health, youth empowerment. Each team had reports. Each program had dashboards. Each region had its own success story. On paper, everything looked important. In the funding committee meeting, the core question was: “Which programs are performing best?” So the team presented: Beneficiaries reached Budget utilization Cost per participant Activity completion rates Good business questions. But no one could commit to a funding shift. Because “best” is descriptive — not decisive. So I interrupted with a harder question: “If your donor cut funding by 30% tomorrow, which two programs would you protect at all costs — and which ones would you pause?” The room went quiet. Now the data had consequences. We reframed the analysis around funding decisions, not reporting metrics: Instead of “How many people did we reach?” We asked: Which programs change outcomes per dollar spent? Which ones absorb cash but stall on impact? Which regions need investment versus redesign? One program reached thousands but produced weak post-program outcomes. Another reached fewer people but permanently shifted income levels. A third had strong stories but no measurable sustainability. Those truths were hidden when the goal was “report performance.” They became obvious when the goal was allocate capital. By the end of the session, leadership wasn’t asking for prettier dashboards. They were asking: “What thresholds trigger extra funding?” “When do we freeze a program?” “What evidence justifies scaling?” The nonprofit didn’t just review programs. They built a decision system for funding. That is the difference between business questions and decision questions in nonprofit leadership. Business questions ask: “What happened?” Decision questions ask: “What do we fund, fix, or stop next?” In African nonprofit environments — where donor trust, community impact, and scarce capital collide — data must do more than report. It must force choice. My work is building decision frameworks that turn nonprofit data into funding action, not just donor reporting. If your organization has impact data but still debates instead of decides, you don’t need more metrics. You need better decision questions.