ImpactOS · How EdZola works
Every project starts where the last one finished.
ImpactOS is the method behind everything EdZola builds: one way of working, a growing body of sector knowledge, and AI tools that always have a person checking. You don't buy it. Every system we build for you runs on it.
The problem
Most nonprofit systems are built from a blank page.
Two programmes in two cities ask the same questions about who they serve and what changed. Usually both pay to answer them from scratch. Sound familiar?
Discovery starts at zero
Weeks of workshops to rediscover what a beneficiary record needs, how donors move through a pipeline, or which leave rules an NGO runs on.
Quality depends on who you get
Every consultant has their own way. One build is careful, the next ships with gaps nobody checked for. You can't tell which until it's live.
What's learned walks out the door
The project ends, the consultant leaves, and everything they figured out stays in their head. The next organisation pays to learn it again.
ImpactOS exists so that doesn't happen to you, or to the next organisation after you.
What it is
Three layers. Each one moves at a different speed.
The method stays steady so the knowledge and tools on top of it can improve quickly without breaking how your project runs.
Method
The same stages, sign-off points and named owners on every engagement, whether it's a small HR system or a year-long programme tracking system.
For you: you always know what stage you're in, who's accountable, and what has to be true before we move on.
Knowledge
Everything we've learned that's worth reusing, written down once and improved on every project.
- MAP framework
- Standard nonprofit data models
- CRM and HR templates
- Sector playbooks
- 80G, FCRA and CSR fields
- Quality standards
For you: your project starts from what's already been tested elsewhere. Workshop time goes on what makes you different.
Tools
AI and automation that do part of the work: drafting blueprints, configuring systems, cleaning data, running voice interviews, checking for security gaps.
- Blueprint drafting
- System configuration
- Data cleaning
- Voice data collection
- Security checks
For you: faster delivery and fewer manual errors, with a named person reviewing every output before it counts.
Think of a car production system. Nobody buys it, and every car on the line is made with it, to the same standard. ImpactOS does that for EdZola's work. Whoever leads your project, it's built the same careful way.
What you get
Six things true of every project we run.
We don't start from zero
Templates, standard data models and playbooks come first. Nobody designs your system from a blank page.
Clear sign-off points
Four points where you agree before we move on, including locking the design before anyone builds.
One named owner
One accountable person at every stage. You never have to wonder who to call.
Checked before you test
Quality checks against our delivery standards and a security scan before your team ever sees it.
Done means adopted
A project closes when your team uses the system on its own, not on the day it launches.
Every project teaches
What we learn goes back into the templates, so the next system, including your next one, is better.
How we work
From the first conversation to a system your team owns.
Five stages, four sign-off points. The same for every engagement.
Listen
We understand how your work really runs and match you with a practitioner who's built it before.
Scope
Clear requirements, an honest estimate, and what we won't do. If a spreadsheet is enough, we'll say so.
✓ Scope agreedDeliver
Design with you, then build, move your data across and train in usable stages.
✓ Design locked✓ Tested
Close
Handover, documentation, and an adoption check. We stay until it's in use.
✓ In useLearn
A retrospective. What worked goes back into ImpactOS for the next organisation.
The method is written down and owned inside EdZola, so it doesn't depend on any one person. If you want to see the stages and sign-off points for your kind of project, ask us on the first call.
Where AI fits
AI drafts. People review, decide and send.
AI handles volume. People handle consequence. Every AI tool in ImpactOS has to meet five rules before it's allowed near your project.
- Has a placeIt attaches to a named stage of the method. No tool runs just because it can.
- Leaves a trailWhat it produces goes back into the record with its source, so you can see where every answer came from.
- Gets checkedA named person reviews its output before it counts, and before it reaches you.
- Has an ownerOne accountable person at EdZola is responsible for it working properly.
- Can be switched offThe manual route still works. Your project never depends on a tool we can't do without.
In practice
What changes in your project.
A typical engagement, the usual way and on ImpactOS.
Every programme tracking system starts from your Theory of Change and our standard data model.
| Stage | Usually | With ImpactOS |
|---|---|---|
| Requirements | Long discovery, a requirements document written from scratch, scope that keeps moving. | Theory of Change, outcomes and indicators first, using the MAP framework. The data model is locked with you before anyone builds. |
| Design | A new data model for every app, with its own naming. | Start from standard records like beneficiary, session and attendance, tested across 60+ projects. Only what's different about you gets designed. |
| Build | Forms and logic built by hand, rewritten on every project. | AI builds forms, reports and workflows in a test environment from the approved blueprint. Our developers review every change. |
| Test | Most bugs found by your team during testing. | Checked against our quality standards and scanned for exposed personal data before you test. Every bug gets a root cause. |
| Adopt | Handover is one training session. | Dashboards used in your own review meetings, and adoption checked before we close. |
No donor system starts from a blank setup. Each one starts from our template and leaves it better.
| Stage | Usually | With ImpactOS |
|---|---|---|
| Requirements | Discovery run from scratch, requirements rewritten every time. | You fill a short requirements template before the first call. The call only covers the gaps. |
| Design | The same donor fields and pipelines rebuilt on every project. | Start from a standard nonprofit donor setup with donor types, pipeline stages, and PAN, 80G and FCRA fields already defined. Only your differences get designed and signed off. |
| Build | Configured click by click. Quality depends on who builds it. | Configured from the approved design in a test environment, with a change log a consultant reviews. |
| Move data | Donor data cleaned by hand in Excel. | AI cleans donor data, removes duplicates and fits it to a standard import template, then a person checks it. |
| Close | Closed at launch. | Closed after an adoption review: are deals moving and receipts going out? |
Your HR system launches from a tested template, so workshop time goes on your policies.
| Stage | Usually | With ImpactOS |
|---|---|---|
| Requirements | Several calls to understand leave types, shifts, approvals and payroll. | You fill one template: staff list, leave policy, holidays, attendance rules, approvals and payroll. The call only covers what's unclear. |
| Design | Set up from meeting notes, differently every time. | A standard nonprofit setup is the base. Only the differences get discussed. |
| Test | Testing depends on who remembers what to check. | Standard tests for a leave request, an attendance cycle and a payroll run, on every project. |
| Adopt | Your HR team is trained once and left to it. | Your own point person runs one full cycle alone before we close. |
Every voice study adds questions, languages and lessons that make the next one cheaper.
| Stage | Usually | With ImpactOS |
|---|---|---|
| Design | Questions written from scratch over weeks. | Start from a question bank for your sector and study type, with standard consent scripts. |
| Pilot | Language and length problems show up halfway through. | A short pilot checks language, length and drop-off before the full run. You approve it first. |
| Report | Analysis starts from raw transcripts. | AI drafts first-pass themes. A researcher checks every claim against what people actually said. |
| Learn | Lessons stay with whoever ran the study. | Questions that worked go back into the bank. Questions that confused people get flagged. |
Built on ImpactOS
One method underneath everything we offer.
Some of our work is run by our people. Some of it you can use on your own. All of it comes from the same method and knowledge.
Services
An EdZola practitioner runs the work with you, following the playbook for that service.
Products
When a tool proves itself across many projects, it graduates into something you can use without us in the loop.
What we share
Much of what's inside is free to read.
The method and its tools are how we deliver. The thinking behind them is open, whether or not you ever work with us.
Bring us the messiest part of your work.
A tracker, a form, a report nobody enjoys making. Someone who's built it before will look at it and tell you what we'd do, and whether you need us at all.