To Proxy, Or Not To Proxy?
Measuring social impact is messy and complicated; with end goals being (generally) so vast that trying to show progress over a limited period of time usually ends in frustration. Imagine your organization’s vision is “creating a world with gender equality.” If you’re reporting annual progress to that long-term outcome (perhaps for funders), you’ll likely rely on proxy metrics.
Proxy metrics are used when you can’t directly and quantitatively measure something (again, think gender equality). You substitute with a proxy like (for our example) “the number of women in leadership roles.”
As useful as they may be, proxy metrics have a downside. In our example above, simply measuring the number of women in leadership roles leaves out a lot of important information related to gender equality. How many of these leadership positions are in strong, thriving organizations? How many of these women were promoted from within, indicating a culture of gender equality within the organization? A proxy doesn’t give you that kind of depth.
That being said, proxy metrics work wonderfully if they are relevant to the incremental goals identified along a strategic pathway. In the same example, there would be a sequence of dependencies like the financial strength and mission impact of the organization (good/mediocre/poor) and whether there are leadership development opportunities for internal staff (true/false). When those dependencies are met, the number of women in leadership roles at the subject organization is a proxy metric you can sink your teeth into.
In all likelihood, you’ll have a mix of real and proxy metrics to measure the impact of your work. The more complex the goal, the more likely you’ll need a proxy. Give the proxy context by having relevant metrics feeding into it. Follow your logic model and always ask the question--does this matter?