00:00:01 Introduction and agenda

Welcome to the webinar. The agenda for today begins with a discussion of various implementation patterns for information management systems, drawing on 15 to 20 years of experience working in the social impact sector. Following the overview of these patterns, the discussion will pivot to how specific features in ActivityInfo have been developed to help deploy these systems. This will include a deep dive into the design philosophy and BeDataDriven's purpose in helping organizations manage their data better. Finally, the session will conclude with a live demonstration of the ActivityInfo tool in action. The format is casual, and attendees are encouraged to submit questions or differing perspectives in the chat.

00:01:42 The purpose of information management systems

Organizations do not implement information management systems or gather data without a specific reason. These systems are often implemented for case management to ensure that high-quality services are delivered to beneficiaries and rights holders. In humanitarian interventions, systems are used to ensure that food, cash, non-food items, and other goods and services reach the most vulnerable populations. Additionally, these systems are crucial for monitoring and evaluation. They help measure programming, verify if an impact is being made, and provide the necessary information to improve programming to better support individuals and communities.

A modern information management system goes far beyond simple data collection or surveys. To truly improve impact, the managed information must be connected with various decision-makers, including operational staff, donors, and headquarters management, to provide actionable insights. This requires the ability to store structured data, implement data governance policies, and provide secure, granular permissions for data access. Furthermore, organizations must be able to analyze the data, ask questions of it, and visualize it meaningfully to generate information products like dashboards and reports. Information systems have always been essential to these efforts, as evidenced by the existence of a 1,500-year-old cuneiform spreadsheet in Egypt, proving that humans have long needed to use information to plan and execute activities.

00:04:40 Evaluating implementation patterns and risk tolerance

The different implementation patterns seen in the sector can be broken down into specific groups. Based on years of experience, it is clear that all of these implementation patterns come with distinct trade-offs, pros, and cons. When evaluating these systems, organizations must consider their own risk tolerance across various dimensions, such as security, scalability, and time to implement. Organizations need to decide if they are comfortable bearing the risks associated with their shortcomings, or if they would be more comfortable having a partner, vendor, software developer, or consultant own that risk. The presenters acknowledge their own perspective as the builders of ActivityInfo, noting that their primary goal is to reduce risk along these different dimensions.

00:06:27 Spreadsheet-centric systems

The first and most prevalent implementation pattern is the spreadsheet-centric system. It is the de facto option; if an organization makes no other decision about how to manage program information, they will default to creating one or more spreadsheets. These spreadsheets are generally integrated with other tools, such as visualization software, business intelligence tools, or open-source data collection tools that feed data into the system. When mapping out dimensions like cost efficiency, ease of use, implementation speed, ease of maintenance, IT dependency, data quality, scalability, security, and flexibility, most customers start with spreadsheets. Functionally, there is no additional cost, as organizations already pay for Google Workspace or Office 365. It is familiar to almost any knowledge worker, making it very easy to get started, and people can build incredibly sophisticated systems this way.

However, there are significant trade-offs. The skills required to design and implement complex spreadsheet systems usually reside with a single person, creating an unmanageable dependency on that individual. Spreadsheets lack robust data validation capabilities, which can lead to poor data quality. They are not secure and only offer binary access—you either share the file or you do not—without any fine-grained permissions. Furthermore, they do not scale technically or organizationally when managing thousands of records across multiple teams.

A prime example is the Cash Consortium in Yemen (CCY), a group of NGOs and the IOM providing household cash to vulnerable populations. They collected data using mobile tools and paper, then moved that data into spreadsheets to merge it. They struggled with slight variations in name spellings and phone number formats, quickly reaching the limits of using a spreadsheet as a database. They also hit quantitative limits, as Excel is not designed to handle over 100,000 rows on a shared drive. Moving this data into a proper database in ActivityInfo significantly sped up their operations. Spreadsheets are a great solution for two people managing a few thousand rows in a single office, but they break down when more people are added, formulas are altered, or data is shared across different locations.

00:12:45 Externally developed custom systems

At the opposite end of the spectrum are custom-developed systems built by external providers. These are software development companies or consulting firms whose expertise is building and hosting web-based software on servers or cloud infrastructure. The clearest benefit of this approach is absolute control. As long as requirements are well-documented, you can pay a vendor to build a highly usable, custom-fit solution that collects excellent quality data.

However, there are major challenges, primarily cost and time. Building production-ready software correctly is complex and requires expensive expertise. While a pilot or proof of concept might be delivered quickly, a production-ready system takes extensive time to test, iterate, and release. While these systems can be secure and scalable, the organization is entirely reliant on the vendor to implement and maintain that security as technology changes. Any updates to maintain functionality, improve security, or extend features as programs change require new development projects and additional costs. This pattern is a good fit only if you have well-defined requirements, a trusted partner, and guaranteed long-term funding.

The "Cone of Uncertainty" from Steve McConnell's book "Software Estimation" highlights the risks of custom development. Building line-of-business software is not like building a house from a standard blueprint; there is massive uncertainty at the beginning regarding user reactions and future program changes. Even after pilots and field tests, requirements will inevitably shift as reality meets the initial plan. Statistics show that 31% of these projects are canceled before delivery, and 53% run nearly twice over budget. For instance, a UN agency budgeted for custom software to manage part of the humanitarian program cycle. Because the humanitarian landscape changed so rapidly, by the end of the year and the exhaustion of their budgeted hours, they only had a proof of concept. The vendor was responsive, but constantly redoing work consumed the budget. While there are great success stories, this approach carries significant risk.

00:19:50 Internally developed custom systems

This pattern is similar to external development, but the software development is done in-house. This is frequently seen in larger international NGOs and UN agencies with complex organizational structures and larger budgets. The pros are similar to external systems, offering maximum control, but with the added benefit that the institutional knowledge gained during the building process stays in-house. However, this is only true if there is low staff turnover.

A deadly risk for this pattern is funding variance. A team of three developers can cost $150,000 to $200,000 a year. If funding decreases or is lost, the organization is stuck. Recently, due to a reset and pressure on funds in the humanitarian sector, leadership has less appetite to invest in in-house technological projects. They question why a humanitarian or development organization should act as a software development firm. Consequently, many are implementing "buy first" policies, preferring the predictable costs and reduced security and hosting risks of commercial systems.

Another massive frustration and risk is the rapidly changing technology landscape. Software is not a machine with a ten-year lifespan. Internal organizational needs evolve, and changing external technology standards can force an organization to move forward even if their software perfectly suits their current needs. For example, a desktop-based humanitarian coordination platform built in the late 90s had to be shut down because the organization could not secure a $50,000 budget for a necessary security update, making it irresponsible for the IT department to keep running it. Similarly, when Google released the Angular 2 toolkit and stopped supporting Angular 1, development teams worldwide scrambled to rebuild their web applications. These changes are hard to anticipate. However, this pattern can work well for organizations with highly specific, consistent missions over long periods. One US organization supporting vulnerable populations has successfully maintained an internal development team for 40 years because their mission is stable, they have a large staff of caseworkers, and they have the budget to support it.

00:28:03 Commercial off-the-shelf (COTS) platforms

Commercial off-the-shelf (COTS) platforms are solutions bought from a software vendor and configured to an organization's needs. Common examples in the social sector include Customer Relationship Management (CRM) tools like Salesforce, and Enterprise Resource Planning (ERP) systems like SAP or Microsoft Dynamics, which are often used for HR, accounting, and finance. These systems are focused on specific functions but are configurable. Organizations favor them because they offer predictable pricing models, managed infrastructure (meaning the organization does not host it or own the security risks), and much faster deployment times since the software already exists. If the software company is successful and trusted, the tools tend to be very good.

The challenges with COTS systems lie in their rigidity. Because they are built for a specific purpose, it can be difficult to adapt your workflow to fit their solution scope. While they can be sophisticated, extending them to your specific needs might require hiring an expert for custom development. Furthermore, you must trust that the vendor's interests align with yours long-term and that they remain a sustainable company. Using a CRM for activity tracking or case management is possible, but it is not done easily or cheaply. For example, an INGO used a sophisticated CRM for activity tracking. While it was secure and scalable, they had to pay a developer to configure it initially and pay again for any subsequent updates. Additionally, the complex user interface required extensive training for their non-technical staff. Organizations must consider if they truly want to use software designed for salespeople to deliver humanitarian aid.

00:33:35 Summary of trade-offs and risk tolerance

All four of these implementation patterns can result in successful, effective information systems. The critical question is where an organization is comfortable assuming the risks associated with these trade-offs. Organizations must ask themselves if they are willing to risk insecure, potentially low-quality data in favor of the low costs and familiarity of spreadsheets. Alternatively, are they willing to risk cost overruns and longer delivery times to outsource custom development to a trusted partner? Do they want to assume the responsibility of internal development, managing their own infrastructure, and controlling security protocols for maximum control? Or, can they work with the limitations of COTS systems to ensure cost predictability, strong security, and scalability? Asking these questions helps identify the best fit for an organization exploring information management systems.

00:35:46 ActivityInfo as a configurable COTS platform

ActivityInfo is defined as a highly configurable COTS platform used to implement information systems for program data in the social impact sector. The platform's origins stem from the founder's early career as a junior program officer with the UN in Kunduz, Afghanistan, and later with UNICEF in the DR Congo. Tasked with tendering contracts for 150 schools and clinics, he had to build internal systems and configure Microsoft Access databases. He experienced firsthand the high costs and slow speeds of developing new custom systems for every new project or changing humanitarian crisis. In emergency responses, there is no time for month-long development sprints when focus shifts from cholera to school rehabilitation. The team realized they needed a configurable system so they wouldn't have to reinvent the wheel every time, allowing them to hit the ground running with new programs and reuse technology across different countries. This led to the creation of ActivityInfo 15 years ago.

ActivityInfo is designed to make configuration as easy as possible to lower the cost of change and reduce time to value. Unlike CRMs that require hiring external consultants, ActivityInfo empowers field staff to manage and evolve the system themselves. Functional tools and proofs of concept can be built in a matter of days or weeks, not months. The platform excels in implementing governance and maintaining data integrity to prevent "garbage in, garbage out." It provides out-of-the-box building blocks like audit logs, highly granular role-based permissions (e.g., restricting a caseworker to viewing data only from a specific province), and complex data quality rules. Even simple features, like defining field types as numbers, months, or geographic points, immediately improve data quality.

The platform also absorbs the risks associated with system maintenance and staff turnover. ActivityInfo manages all infrastructure security, updates, and versions transparently. When organizational turnover occurs, the support team steps in to work with the organization, fill gaps, and ensure new staff are trained. The system is highly scalable; organizations often start using it for a specific project's M&E and easily expand its use geographically or thematically across different teams. Security is taken very seriously, with an audited and certified Information Security Management program, out-of-the-box single sign-on, and dark web scanning to flag compromised accounts. Finally, the platform is highly flexible, making no assumptions about the data being stored, and ensures data portability so information can be easily extracted or integrated with other ERP or CRM systems.

00:49:00 Demonstration of ActivityInfo

The demonstration highlights an ActivityInfo database used for a specific program, utilizing a template designed for primary data collection. It features a registry of farms and individuals that acts as a single source of truth. The web interface includes a robust data cleaning tool that allows users to copy, paste, and make changes directly within the interface without needing to export the data. It offers multiple views, including map visualizations to pinpoint work locations, which can be tailored for different stakeholders.

Data is collected through integrated mobile apps for Android and iOS, which are recommended over third-party tools. The app works both online and offline. Users can download the entire database to their device for two-way synchronization, allowing them not only to send new data but also to review and edit previously collected records. The app automatically flags duplicates, blocking users from registering a farm with an identical name and postcode.

Once collected, the data connects directly to reports and dashboards that display KPIs, maps, and real-time indicator tracking calculated from the primary data. Crucially, ActivityInfo is a relational database. This allows users to link the registry of individuals to specific activities. When adding a record, users can look up participants and link the same person to multiple activities or sessions over time. This relational capability allows organizations to follow a person across time to truly measure impact, such as pulling data from multiple surveys to calculate the percentage increase in crop yield for specific farms—a complex task that is very difficult to achieve with spreadsheets.

00:54:30 Q&A and conclusion

An audience poll conducted during the session revealed that nearly 50% of the attendees use spreadsheet-based tools, while about a fifth use custom systems developed by external parties. During the Q&A, a participant asked about French language support. The presenters confirmed that while live subtitles are not always available, subtitles in French and Spanish are provided for all webinar recordings.

Another participant asked about the system's backend architecture and whether it relies on Microsoft. The founder explained that because ActivityInfo is designed to run offline in web browsers and on mobile phones, they had to build their own custom database engine. For the hosted version, the backend storage utilizes Google Cloud Datastore. However, there is also a self-managed version available for download that uses SQLite for storage. Within the web browser itself, the system runs on IndexedDB. Because of these different layers, users do not access the storage layer directly; instead, they use an API to access their data. The webinar concluded with the presenters thanking the attendees for joining.