How Advocacy Campaigns Measure Reach

Published by Instant Press on behalf of Fight For A Living Wage.

Advocacy groups measure reach with a stack of counts that sit at different distances from a human being actually paying attention, and the honest ones are almost always the smallest. Impressions sit furthest away from attention. Completed volunteer shifts, replies to an email, and survey-measured awareness sit closest. One rule holds across the whole stack: prefer a figure you can rebuild from your own records over a figure a platform hands you.

That preference costs an organization bragging rights and buys it the ability to tell whether a campaign worked. What follows is the mechanism behind each common metric, then illustrative arithmetic showing how a large headline number shrinks into a small real one.

What an impression actually records

An impression records that a piece of content was served into a space where a person could have seen it. Platforms count a video view after a few seconds of playback, sometimes with sound off and the window already scrolling past. An outdoor advertising impression counts the vehicles that drove a route. No impression count confirms that a person read a word, held the message, or could restate it an hour later.

The distance between delivery and attention grows with the size of the buy. Broad targeting reaches people with no stake in the issue, and those people scroll. If your organization reports impressions at all, print the measurement window and the platform definition next to the number. A view counted at two seconds and a view counted at fifteen seconds describe different human experiences, and averaging them produces a figure nobody can interpret.

Earned media mentions and share of voice

Earned media counts the times journalists wrote about the work without a placement fee. It is a better signal than paid impressions because an editor had to decide the story was worth space. Share of voice takes the same mentions and compares them against other organizations discussing the same issue over the same window, which tells you whether you are shaping the conversation or standing next to it.

Both counts flatten things that matter. A mention in a regional trade newsletter and a mention in a national broadcast each register as one. Weight mentions by the outlet’s audited circulation or audience where that figure exists, and separate the cases where staff were quoted as a source from the cases where the organization appeared in a list. Those two kinds of mention produce different downstream traffic, and mixing them hides the difference.

Petition signatures as a weak proxy

A signature measures one click on one form. The count rises with the ease of the form, the width of the aggregator network carrying it, and the emotional pitch of the headline. Duplicate submissions, disposable addresses, and signers who arrived through a partner list and will never recognize your name all sit inside the total. A campaign can double its signature count by shortening the form, which changes the number without changing support.

The version of the metric worth tracking is the share of signers who later did a second thing: opened a follow-up email, attended an event, gave a phone number that connected. That ratio is small, stable, and hard to inflate, which makes it useful.

Volunteer conversion, the number worth defending

A completed volunteer shift leaves a record with a timestamp, a location or a phone bank log, and a coordinator who can confirm it. Report sign-ups and completed shifts as two separate figures, because the ratio between them measures how well the organization runs its own onboarding. A group that recruits a thousand sign-ups and seats eighty people has a scheduling problem, and a single blended number hides it.

Email and list engagement

Open rates stopped carrying much information once mail privacy features began prefetching images on behalf of recipients. Clicks, replies, and unsubscribes still reflect decisions a person made. The figure worth putting on a dashboard is the share of the list that took any recorded action in the last ninety days, reported alongside total list size. A list of 50,000 addresses with 3 percent recent activity is a smaller asset than a list of 6,000 with 20 percent, and only the second framing tells you that.

Survey-measured issue awareness

Asking a representative sample an unprompted question comes closest to measuring whether a message landed. The method is slow and expensive, the samples are small, and the margins of error are wide enough that a single wave rarely proves movement. Report the margin of error and the field dates every time, and compare waves rather than reading one in isolation.

Worked example: illustrative arithmetic

The numbers below are constructed to show the shape of a funnel. They measure no actual campaign, and each rate is a plain round figure chosen for arithmetic rather than drawn from any dataset.

Start with a stated 2,000,000 impressions.

• Apply a 1 percent engagement rate: 20,000 engagements.
• Assume a quarter of those engagements are link clicks rather than likes: 5,000 landing page sessions.
• Assume 60 percent of sessions end in under ten seconds: 2,000 sessions long enough for reading.
• Assume 8 percent of those readers give an email address: 160 new subscribers.
• Assume 10 percent of new subscribers complete a first action: 16 people.

Two million at the top, sixteen at the bottom, a ratio of 125,000 to one. Nothing in that chain is pessimistic. Every rate is ordinary, and the collapse comes from multiplying ordinary rates five times in a row. The useful part is that each line names a place to intervene. Moving the eight percent sign-up rate to twelve percent adds eighty subscribers for the cost of a better page, which is a cheaper gain than buying another million impressions.

Why a smaller verified number is worth more

A funder can audit sixteen completed actions. Those sixteen people can be called again, surveyed, and asked to bring someone. A two million impression figure supports none of that, and when the campaign underperforms the large number offers no clue about which step failed. Publishing the small figure also protects the organization the next time a reporter asks how the count was produced.

Groups working on affordability have a particular reason to hold this line, because the underlying facts are already large. The federal minimum wage has sat at $7.25 an hour since 2009, according to the U.S. Department of Labor. Median household income ran about $80,000 in 2023, according to the U.S. Census Bureau, while the median home sale price sat near $400,000 to $420,000 in 2024 in National Association of Realtors and Census figures, roughly five times income against about three times in the 1980s. The MIT Living Wage Calculator sets county-level costs beside prevailing pay, and the Bureau of Labor Statistics publishes the wage and price series underneath both. A campaign citing those sources and showing its arithmetic has no need of an inflated audience figure.

Publishing the underlying data belongs to the same discipline. Fight For A Living Wage, a nonpartisan grassroots 501(c)(3), maintains a public data hub and a research library on its site, so a reader can trace a figure back to its primary source instead of trusting a summary.

A short standard to adopt

Write a one-page definitions sheet naming, for every metric the organization publishes, the source system, the counting rule, and the date range. Put sign-ups and completions in separate columns. Keep impressions in a clearly labeled section away from verified counts so nobody adds them together. Then hold the same definitions for three years, because a reach figure only means something once you can compare it with the one before it.