Every scored diagnostic produces a number, and almost none of them will show you how. This is the whole method: the benefits you can rate, the three questions asked about each one, the points behind every answer, the weighting, the tier thresholds, and an honest account of what the result does not tell you.
The test offers 16 common association benefits and lets you add your own. You choose at least four and up to twelve. Eight gives a representative portfolio without the session running long.
Each benefit you select is rated on the same three questions. Every answer carries points, shown below, and the three are weighted rather than averaged.
If you stopped offering “your benefit” tomorrow, where would most members turn?
| Answer | Points |
|---|---|
| A free AI tool or web search would fill the gap | 0 |
| A competitor, vendor, or free community offers something close | 1 |
| They could patch something together, but would lose quality or trust | 2 |
| Nowhere. There is no real substitute | 3 |
Carries the most because it is the only one that measures exposure directly. If a member would simply ask a free tool instead, the other two questions cannot rescue the benefit. Everything else in this framework is a reason a benefit might survive; this is the question of whether it is under threat at all.
What does “your benefit” primarily run on?
| Answer | Points |
|---|---|
| Publicly available information that we package and distribute | 0 |
| Our staff's general expertise and curation | 1 |
| Data, research, or content that only we hold | 2 |
| Authority only we hold: standards, accreditation, certification, a seat at the table | 3 |
What a benefit runs on determines whether it can be copied. Public information that an association repackages is reproducible by anyone with a search box. Proprietary data, accreditation authority, or a seat at a regulatory table are not, because the barrier is access rather than effort.
How much of the value of “your benefit” comes from people: relationships, judgment calls, or recognition?
| Answer | Points |
|---|---|
| Very little. Members mostly consume content | 0 |
| Some. There is a human layer, but content does the work | 1.5 |
| It is the point. Members could not get this alone or from a machine | 3 |
Relationships, judgment calls and recognition are the categories a generative system cannot supply, because their value depends on a person being accountable for them. Weighted equally with foundation because in practice they protect different benefits: data protects research and benchmarking, people protect community, mentorship and credentialing.
The maximum points on any single question is 3, which is why the weighted total is divided by 3 before scaling to 100. Questions with 4 answers and questions with fewer are therefore comparable.
| Tier | Benefit score | Verdict |
|---|---|---|
| Commoditized by AI | below 38 | Stop competing here |
| Contestable | 38 to 61 | Win it or lose it |
| AI-Resistant | 62 and above | Invest here |
| Stage | Portfolio score | What it means |
|---|---|---|
| Exposed | 0 and above | Most of what your members pay for, a general AI tool can now approximate. This is urgent but fixable: the goal is not to defend everything, it is to decide deliberately what to concede, what to contest, and where to rebuild the core. |
| Contested | 40 and above | Your portfolio is split. A meaningful share of your member value is still defensible, but a large contestable middle will be won or lost in the next few years based on whether your data, curation, and governance are strong enough to compete. |
| Defensible | 62 and above | The core of your member value sits where AI cannot follow: judgment, standards, community, and accountability. The work now is to keep investing there, and to stop subsidizing the benefits AI has already commoditized. |
We want to be explicit about this, because our AI readiness framework is different. There, the dimension weights are argued from published research. Here they are not. The Member Value Defensibility Test is an original CNAI framework and no study validates these three factors or the 40/30/30 split. They represent our reasoning about what determines exposure, published so you can disagree with them specifically rather than generally.
Every answer is a staff view of what members value and where they would go. Association executives and their members frequently disagree about that, and the gap is often the whole story. If you want a sharper reading, have two or three colleagues take it separately, and better still, ask some members the substitutability question directly.
A benefit landing in the commoditized tier does not mean cancel it tomorrow. Some commoditized benefits are cheap to run and members still expect them. The tier tells you where competing harder is unlikely to pay, which is a budget and attention question, not an existence question.
Nothing here has been tested against renewal data. A portfolio that scores well is one whose value is structurally harder for AI to replace. Whether that shows up in dues revenue depends on execution, pricing, competition and a dozen things this instrument does not look at.
If you think a weight is wrong, a question is poorly framed, or a benefit is missing from the list, we want to hear it. Write to us through the contact page. Changes to the instrument will be recorded on this page rather than made quietly.
Instrument version 1. This page is generated directly from the live test, so the questions and point values above are always the ones being used.
Center for Nonprofit