# Building a customer advocacy program

> Stand up a SaaS advocacy program in 90 days: finding advocates, the ask ladder, rewards that pass procurement, tooling, and advocacy influenced pipeline.

Source: https://saas-marketing.net/playbooks/customer-advocacy-program/
Topic: SaaS Customer Marketing
Type: playbook
Published: 2026-09-11
Last updated: 2026-09-11
Publisher: SaaS Marketing (saas-marketing.net)
License: CC BY 4.0. Quote or republish with attribution and a link to https://saas-marketing.net/playbooks/customer-advocacy-program/

## Short answer

A customer advocacy program is a structured system for identifying happy customers and matching them to public acts of support: reviews, references, case studies, webinars, speaking slots and community answers. It works when each ask returns something the advocate wants, usually visibility, access or peer connection. Build it in 90 days by sourcing advocates from health score and product usage data, laddering asks by relationship depth, and capping requests per account.

## Key takeaways

- Source advocates from health score, NPS promoters and product usage data, not from a CSM's memory of who was friendly.
- Ladder asks from a 10 minute review up to a conference keynote, and never skip rungs on a new advocate.
- Cap asks at four per account per year or your best advocates quietly stop replying by month eight.
- Points and badges burn out inside two quarters because they reward the program, not the person.
- Reference requests from sales need an intake queue with a 48 hour SLA or reps will go around you.
- Advocacy influenced deals typically close at a higher rate, so measure win rate lift against a matched control.

---

Advocacy programs fail in a predictable way. Someone buys a platform, sets up a points system, and sends 300 customers an invitation to earn badges. Two quarters later there are 40 registered advocates, eleven of whom have done anything, and the program owner is quietly reassigned. The failure isn't the tooling. It's that the program started with a mechanic instead of a value exchange, and customers worked out within a week that the exchange was one-directional.

## Who your advocates actually are, and how to find them

Your advocates are already visible in your data. You do not need to ask anyone who they like.

Build the list from three signals joined together: health score above your internal threshold, an NPS response of 9 or 10 inside the last two quarters, and product usage in the top quartile for their segment. Then subtract anyone with an open support escalation, a renewal risk flag, or a contract under 90 days from expiry. What comes back is usually 3 to 8 percent of the customer base, which for a 2,000 customer company is 60 to 160 names. That is more than enough to start.

Add a fourth signal that most teams ignore: community and social behaviour. People who answer questions in your community forum, who have posted about the product unprompted, or who show up in your Slack Connect channels with product opinions are pre-qualified. They've already advocated. You're just formalising it.

Asking CSMs "who would do a case study" produces the same eight names every quarter, all of them already exhausted, all of them from the accounts the CSM enjoys talking to. It systematically misses quiet power users and every account in a segment the CSM covers lightly. Query the data first, then have CSMs veto rather than nominate.

**3 to 8%** Share of a SaaS customer base that qualifies as advocate-ready on health, NPS and usage

## The ask ladder, and what the advocate gets at each rung

The single design decision that determines whether your program survives is the ladder. Start small, prove the exchange works, escalate. Never open a relationship with the biggest ask.

The value returned has to be real and specific. A reference call is worth doing because the advocate gets to talk to a peer solving the same problem, which is genuinely useful to them, and because you introduce them properly and follow up. A case study is worth doing because you actually distribute it, tag them, and send them the traffic numbers afterward. If you don't do those things, the exchange is fake and the advocate learns it on the first ask.

Notice that community answers sit outside the ladder. That's deliberate. People who help others in your community are doing it for status and identity reasons, and pulling them onto a transactional ladder can break the thing that motivated them. Atlassian Community Leaders and Salesforce Trailblazers both work this way: recognition, access and peer standing, with no points balance in sight.

## Rewards that survive enterprise procurement

This is where programs get quietly killed by a customer's legal team, and where most advocacy platform default settings will get your advocate in trouble.

Most public companies cap gifts at 25 to 100 dollars. Financial services, healthcare and government customers frequently accept nothing of monetary value at all. A points system that accrues toward a 500 dollar reward is unusable for a large slice of an enterprise customer base, and worse, it puts the advocate in the position of explaining a gift log entry to their compliance team.

Rewards that work across almost all policies:

- Early access to unreleased features, with a direct line to the product manager
- A named seat on a customer advisory board, which is a genuine line on a résumé
- A conference ticket and travel, tied to their own speaking slot so it's professional development rather than a gift
- A charitable donation in their name to an organisation they choose
- Distribution: your audience seeing their name and their work

Always include a decline-the-gift path that still records the contribution. And write the reward policy down before the first ask, because the moment you improvise, someone offers a customer an Apple Watch and you spend a week with legal.

Gamified advocacy programs peak around month four and collapse by month eight. The mechanic rewards the program's goals, not the person's, and B2B professionals have no reason to accumulate points from a vendor. Ladder by relationship depth and recognition instead. Notion's ambassador structure and Figma's community programs both lean on status and access rather than a leaderboard.

## Program operations: the intake queue is the whole job

Half of running advocacy is routing requests from sales without letting sales trample the advocates. If you don't build a queue, reps will email customers directly, and your best reference will get three unrelated requests in one week.

**Reference request operations**

Tooling in the first 90 days should be boring: a spreadsheet or Airtable base for the advocate roster, a Slack channel for intake, and custom fields on the CRM opportunity object. Buy a dedicated advocacy platform when you pass roughly 150 active advocates or when one person can no longer route the volume. HubSpot's customer marketing team ran a substantial reference operation on CRM fields and process discipline long before any of it was automated, which is the right order.

## Measuring advocacy influenced pipeline without fooling yourself

Two measurements, and they answer different questions.

The first is influence volume: pipeline where an advocacy touch is logged on the opportunity. Reference call, case study page view attributed to the account, peer conversation at an event. This is easy to produce and easy to overstate, because references are disproportionately requested on deals that were already going to close. A rep asks for a reference when the deal is real.

The second is the one that matters: win rate lift against a matched control. Take deals with an advocacy touch and compare them to deals matched on segment, deal size, stage entered and quarter, without a touch. If the treated group wins meaningfully more often, or closes faster, you have something. If the two look identical, your references are a comfort blanket and the program should be redirected toward review volume and content instead.

| Metric | How to measure it | What good looks like | Honest caveat |
| --- | --- | --- | --- |
| Advocacy influenced pipeline | Opportunities with a logged advocacy touch | 15 to 30 percent of enterprise pipeline | Inflated by selection bias toward strong deals |
| Win rate lift | Treated versus matched control | 5 to 15 point lift | Needs 100+ deals per group to be meaningful |
| Review volume and rating | G2 and Capterra dashboards | 20+ reviews per quarter, rating above 4.4 | Rating moves slowly, volume moves inbound faster |
| Advocate retention | Advocates still active after 12 months | Above 60 percent | Falling retention is the earliest failure signal |
| Asks per account | CRM counter | Under 4 per year | The number the program owner defends |

The [customer marketing ROI calculator](/calculators/customer-marketing-roi/) handles the arithmetic on the influence side. The broader measurement frame sits in [SaaS customer marketing](/saas-customer-marketing/), and if the program stalls, the diagnostic patterns are in [why customer marketing programs fail](/guides/why-customer-marketing-fails/).

Advocate retention. Programs that are extracting more than they return show it here first, before pipeline influence moves, and usually two quarters before anyone admits the program is in trouble. If you're losing more than 40 percent of advocates a year, your ladder is broken or your cap is not enforced.

## The 90 day build

**Days 1 to 30**

**Days 31 to 60**

**Days 61 to 90**

Feeding the program properly means having a structured way to hear from customers in the first place, which is what a [voice of customer program](/guides/voice-of-customer-program/) provides, and advocates who came in through good onboarding convert at a much higher rate than ones recruited cold, so [customer onboarding marketing](/guides/saas-customer-onboarding-marketing/) is upstream of all of this. Definitions and scope are in [what is customer advocacy](/glossary/customer-advocacy/), worked examples in [SaaS customer marketing examples](/examples/saas-customer-marketing-examples/), and if the program grows into a peer network rather than a reference pool, [SaaS customer community strategy](/guides/saas-customer-community-strategy/) is the next build.

## What to do next

Run the advocate query this week. Not the CSM survey, the query. Then pick 20 names and make exactly one small ask of each, with a real return attached, and see what your response rate is. That number tells you more about whether an advocacy program will work at your company than any amount of planning.

If you get above 30 percent response, build the ladder and the intake queue and go. If you get single digits, the problem is upstream in product satisfaction, and an advocacy program will make it worse by exposing it. Fix that first. A structured version of the full build is in the [customer marketing sprint](/courses/saas-customer-marketing-sprint/).

## Frequently asked questions

### How do you find customers willing to advocate?

Combine three data sources: health score above your threshold, NPS score of 9 or 10 in the last two quarters, and product usage in the top quartile of their segment. Then filter for accounts with no open support escalation and no renewal risk flag. That list is usually 3 to 8 percent of your customer base and it is far better than asking CSMs who they like.

### What rewards can enterprise customers actually accept?

Most public companies cap gifts at 25 to 100 dollars and many financial services and government customers accept nothing at all. Safe rewards are non-monetary: early access to features, a product roadmap session, a named seat on an advisory board, a conference ticket tied to their speaking slot, or a donation to a charity they pick. Always offer a decline-the-gift option.

### How long does it take to build an advocacy program?

Ninety days to a working program with 20 to 40 active advocates, a live ask ladder and a reference intake queue. Six to nine months before it produces measurable pipeline influence. Teams that promise pipeline in the first quarter end up burning their best customers on rushed asks and have nothing left when the program matters.

### Should we use an advocacy platform?

Not in the first 90 days. A spreadsheet, a shared Slack channel and your CRM will run 40 advocates fine. Buy a platform when you pass roughly 150 active advocates or when reference request volume exceeds what one person can route. Buying early means you configure a system before you know what your program does.

### How do you measure advocacy program ROI?

Track advocacy influenced pipeline, which is opportunities where a reference call, case study view or peer conversation is logged in the deal record, and compare win rate and cycle length against matched deals without advocacy touches. Also count review volume and rating movement on G2 and Capterra, since those affect inbound independently of any single deal.

### How many asks per advocate is too many?

Four per account per year is the practical ceiling, and no more than two of those should be time-intensive like a webinar or conference talk. Past that, response rates fall sharply and you start hearing the phrase 'we are becoming your marketing department'. Track asks per account in your CRM so nobody can exceed the cap by accident.