customer champion program

How to Build a Customer Champion Program: Steps and Example (2027)

A 5-step method to identify customer champions, score them on success, and measure network value instead of LTV. Includes a 60-account worked example.

The customers who love your product are usually the ones nobody talks to. CS time goes to the accounts that complain, and big logos get attention because they are big. The accounts that are expanding usage and referring colleagues sit in the CRM as one more row. This guide from the BuildBetter team lays out a five-step method for finding those customers, learning from them, and turning them into a source of growth and roadmap direction. It closes with a worked example of a 60-account company picking its first eight champions.

What Is a Customer Champion Program?

A customer champion program is a structured way to identify your most successful customers (roughly the top 10%), learn what makes them successful, give them what accelerates that success, and make it easy for them to bring others. You then measure them on network value instead of their own revenue alone.

A champion is a specific kind of customer. A champion is a customer who:

Succeeds measurably with your product
Actively expands usage through more seats, features, or integrations
Brings others in, including colleagues, peer companies, and former coworkers
Defends you when things go wrong, internally and sometimes publicly

Spencer Shulem's book Customer-Led Development puts it in one line: “a champion builds their career on your product.”

What a champion is not

An NPS promoter. A 9 or 10 on a survey measures stated intent. Plenty of promoters are satisfied but passive and will never refer anyone.
A long-tenure account. Some customers stay out of inertia or because switching is painful.
The loudest account. Frequent contact usually means high maintenance, not high success.

Champions are worth the effort because they compound. They grow their own accounts and pull new people in, so one champion can be worth more than many regular customers combined.

Why Champion Programs Are Hard (and What Usually Breaks)

Most B2B companies under-invest in their best customers because attention follows noise, not success. Standardized onboarding, generic check-ins, and CS time split evenly across accounts mean your healthiest accounts get the least deliberate investment. Several other problems pile on top of that.

Three false positives look like champions

The loudest customer. Lots of requests and lots of calls often add up to a high-maintenance account and nothing more.
The big logo. Size hides shallow adoption. The book's example is 200 seats at 15% adoption, which is a churn risk, not a champion.
The early adopter. A day-one customer may be held in place by switching costs rather than loyalty.

Valuation is wrong by default

Revenue-based customer lifetime value ranks a $100K account that refers nobody above a $50K account that brought in five customers. Academic work backs this up: Kumar, Petersen and Leone found that the customers with the highest lifetime value were frequently not the customers with the highest referral value.

Referrals go unrecorded

Nobody writes down who referred whom. Without that record, the referral tree that would reveal your champions never gets built.

Signals are scattered

Adoption data lives in product analytics, seat growth in billing, referral mentions in sales calls, and community activity in Slack. No single person sees all of it.

Champions move

B2B relationships follow people. A 2021 study in the Review of Accounting Studies found that supplier relationships ended more often when a connected executive left the customer company. Programs that don't track job changes lose champions without noticing.

The Method: The Champion Expansion Play in 5 Steps

The Champion Expansion Play is a five-step method: identify, document, build, enable the network, and measure network value. It comes from Chapter 17 of Customer-Led Development: How to Build What Your Customers Want When AI Can Build Anything by Spencer Shulem. Everything below runs on a shared spreadsheet, your billing export, and a calendar. It needs no budget and no new tool.

Step 1: Identify your top 10% by success, not size

Pick four to six signals you can pull today from data you already have. Weight them for your product, score every account, rank them, and treat the top 10% as champions or potential champions. Refresh the scores every quarter.

Category	Signal	Where to find it today	Suggested weight
Success	Feature adoption above 80%	Product analytics, admin dashboard	High
Success	Usage expanding month over month	Product analytics	High
Success	Multiple users or teams	Admin dashboard, billing	Medium
Success	Proactive learning (webinars, docs, advanced questions)	Call notes, support history	Low
Expansion	Adding seats unprompted	Billing export	High
Expansion	Feature requests from highly engaged users	CRM notes, call notes	Medium
Expansion	Integrations with other tools	Admin dashboard	Medium
Expansion	Product described as “mission critical”	Call notes, CRM notes	Medium
Network	Unprompted referrals	CRM “referred by” and “how did you hear about us” fields	High
Network	Public mentions	Social, Slack communities	Low
Network	Willingness to do a case study	CRM notes	Low
Network	Community activity	Slack, community forum	Low

The evidence usually already exists:

Billing exports for seat growth
Product analytics or admin dashboards for adoption breadth
CRM notes and “how did you hear about us” fields for referrals
Support history for low support burden
Call notes and Slack for unprompted praise and introductions

Then validate. Run the ranked list past CS and ask, “Do you agree these are champions?” CS will catch false positives the data misses, such as a heavy-usage account quietly evaluating other vendors, and false negatives, such as a quiet referrer. Remove the loud, the big-but-shallow, and the trapped early adopters. The rule: score for success, not for size or tenure.

One data fix pays off more than any other: add a required “How did you hear about us?” field and a CRM field for “referred by (contact + account).” Start recording today.

Step 2: Document their exact use cases

Interview your top champions. The book suggests up to the top 20. Ask:

How do you use the product day to day?
What did your workflow look like before?
What would make you stop using us?
What feature could you not live without?

Then map each champion's path: the first aha moment, how long it took to get there, what almost made them churn, and what made them stick. Synthesize the patterns that separate champions from struggling customers. That synthesis becomes your success playbook for onboarding and CS.

Step 3: Build features that accelerate their success

Prioritize the one or two changes that shorten champions' path to value or remove their specific friction, ahead of features aimed at the average customer. Ship to champions first. Early access earns loyalty and validates the feature before broad rollout.

Step 4: Make it easy for them to bring their network

Use one simple referral mechanism. The book is explicit: not a complex program with tiers and points. Give champions easy colleague invites and forwardable content. Give credit through public recognition and a personal thank-you. Arm them with case studies, ROI data, and talking points for their boss.

What champions get	What you ask
Direct lines instead of tickets	Interviews
Early access	Beta feedback
Influence on the roadmap	Reference calls
Recognition	Referrals
Assets that make them look good internally	Occasional public advocacy

Always ask less than you give.

Step 5: Measure network value, not individual LTV

Track direct referrals, indirect referrals (who their referrals brought), influenced deals such as reference calls, and public advocacy. The book's example shows how much this changes the ranking:

Account	Own ARR	Referrals	Referral ARR	Network value
Acme	$100K	0	$0	$100K
TechCo	$50K	5	$175K	$225K
StartupX	$20K	12	$180K	$200K

Referred customers also tend to be better customers. Schmitt, Skiera and Van den Bulte (Journal of Marketing, 2011) found referred customers at a German bank were about 16% more valuable and roughly 18% less likely to churn. That is B2C banking data, so treat it as directional for B2B SaaS.

Feed champions into the roadmap. Tag every feature request from a champion account. When a request appears across several champion interviews, it becomes a Step 3 candidate. Champion friction outranks friction measured by request volume.

Follow champions when they change jobs. Reach out at their new company, and check on the account they left behind before it drifts. UserGems, a vendor that sells job-change tracking, reports 114% higher win rates on deals involving someone who had used, bought, or evaluated the product before. Treat it as a vendor's number, but the direction matches the academic research on executive departures.

The first 90 days

Phase	Activities	Output
Month 1: Identify	Define success criteria, build the scoring, score every account, validate with CS	Ranked champion list
Month 2: Document	Interview up to the top 20, synthesize patterns	Champion success playbook
Month 3: Enable	Prioritize 1–2 accelerating features, launch the referral mechanism and recognition	Active program
Ongoing	Track metrics weekly, refresh identification quarterly	Current list and network-value data

Optional: the champion success manager role

	Traditional CSM	Champion success manager
Accounts	50–100	15–25 champions
Mode	Reactive	Proactive
Playbooks	Generic	Customized success plans
Measured on	Retention	Expansion plus network value

A small team can give this role to one person part-time.

Worked Example: A 60-Account Company Picks Its First 8 Champions

This is an illustrative scenario, not a customer case study. The company is a B2B SaaS business with 60 paying accounts, one CS lead, one PM, and no budget for the program.

The scoring setup

The team scored every account out of 100:

Feature adoption breadth: 30 (admin dashboard)
Month-over-month usage growth: 25 (admin dashboard)
Unprompted seat additions: 20 (billing export)
Referrals or intros on record: 15 (CRM notes)
Low support burden: 10 (support history)

The spreadsheet had one row per account and one column per signal, plus ARR from billing and a free-text “CS notes” column.

Why 8, not 6

A strict top 10% of 60 is six accounts. The team took the top six by score, then added the accounts ranked 7th and 9th because CS flagged both as having referred colleagues. Eight of 60 is about 13%, inside the book's healthy 5–20% identification band.

Who was cut, and why

The largest account, 120 seats at under 20% adoption, scored low. It was reclassified as a churn risk and handled by standard CS.
The account with the most feature requests had a high support load and flat usage. Loud, not a champion.
A day-one customer with flat usage was likely staying because of switching costs, so it was excluded.

What each of the 8 did in Q1

Two did reference calls on open deals.
Two each introduced one peer company.
One joined an early-access beta for the champion-requested feature.
One added a second team after a use-case session.
One shared a workflow that became the onboarding playbook.
One did nothing beyond the interview, which is normal.

The roadmap decision

Five of eight interviews named the same friction point: a manual export step. The PM moved the fix ahead of a feature requested by more accounts overall, because those requests came mostly from non-champions.

End-of-quarter network-value math

Take one champion, a hypothetical account called Northwind, at $30K ARR. Its referral tree, using numbers made up for illustration:

Northwind ($30K own ARR)
- Fernway ($24K), referred two years ago
  - Halcyon ($15K), referred by Fernway
- Brightline ($18K), introduced in Q1 and closed at quarter end

Individual value: $30K ARR. Network value: $30K + $24K + $15K + $18K = $87K, nearly three times its own revenue. The cut 120-seat account brings in about $54K and has no referrals, so its network value is $54K.

What changed for Q2

The team re-scored all 60 accounts.
One champion whose usage declined was dropped, and one new account was added.
Referrals predicted more than seat growth did, so the referral weight went from 15 to 20 and seat additions dropped from 20 to 15.

Common Mistakes When Building a Champion Program

Most champion programs fail on selection, not execution. These are the mistakes that show up most often:

Using NPS as the champion list. In a telecom sample, Kumar, Petersen and Leone (Harvard Business Review, 2007) found roughly 68% of customers said they intended to refer, about a third actually did, and only around 14% of those referrals became customers. Keiningham and colleagues (Journal of Marketing, 2007) found NPS no better at predicting growth than other satisfaction metrics. Score behavior, not survey answers.
Choosing by logo or tenure. Big-but-shallow accounts and inertia accounts dilute the program and eat high-touch time.
Building a points-and-tiers referral scheme. Complexity kills participation. One simple mechanism plus a personal thank-you beats gamification.
Treating champions as a marketing asset. Asking for case studies and reviews before you've removed their friction turns advocacy into a chore.
Generic check-ins. “Just touching base” calls waste champion time. Every interaction should give them something.
Setting criteria too loose or too tight. Below 5% identification signals a product success problem. Above 20% means your criteria are too loose.
Never refreshing the list. Champions emerge and leave. A list scored once in January is wrong by summer.
Losing champions when they change jobs. Without a process, nobody follows them to their new company or re-secures the account they left.

Some of these failures are organizational, and no tool fixes them. Software does not decide to stop gamifying referrals, does not make check-ins worth a champion's time, and does not give a person ownership of champion relationships. Those are choices your team has to make.

When You Need Tooling (and Which Tools)

A shared spreadsheet comfortably handles a first champion program, up to the 15–25 champions one person can manage, as long as the signals come from structured sources like billing and admin dashboards. Manual tracking breaks in two places.

The first is when network and success signals are buried in hundreds of recorded calls, support tickets, and Slack threads: unprompted referrals, “we rolled this out to another team,” “I told my old boss about you.” No one can re-read that volume every quarter, so champions get missed and the list goes stale. The second is when champion feature requests have to be traced from conversation to roadmap item across several PMs, and spreadsheet tagging stops being consistent.

Tool	Best for	Surfaces signals from conversations?	Limitation
BuildBetter	Champion signals buried in calls, tickets, and Slack	Yes	Not a referral-payout or rewards system
Shared spreadsheet	First program, under ~25 champions, structured data	No	Manual entry, misses conversational signals
Airtable	Linked records for the referral tree	No	Manual signal entry
Coda	List, interview notes, and playbook in one doc	No	Relies on someone feeding it data

1. BuildBetter — Best for Champion Signals Hidden in Conversations

BuildBetter finds champion signals inside the conversations where they actually happen. It unifies call recordings, Slack, support tickets, surveys, and CRM data through 100+ integrations, including Zoom, Slack, Jira, Salesforce, Zendesk, HubSpot, and Intercom. Signals pulls referrals, expansion talk, and praise out of each conversation with severity and business impact attached, and Tracked Objects links champion requests to evidence and closes the loop when you ship. Champion requests become PRDs through Documents and Jira or Linear issues through Tickets, so the Step 3 handoff carries full context.

Because it combines internal team activity with external customer feedback, it also catches the signals CS mentions in Slack but never logs in the CRM. Honest caveat: BuildBetter is not a referral-payout or advocacy-rewards system, and teams running enterprise-scale survey distribution may want a dedicated survey platform alongside it.

Other Tools for Running a Champion Program

2. A shared spreadsheet (Google Sheets or Excel)

Best for a first program or under roughly 25 champions with structured data. It is free and transparent, but it cannot surface signals that live in conversations.

3. Airtable

Best when you want linked records for the referral tree (champion to referred account to their referrals) and simple views for CS. It still requires manual signal entry.

4. Coda

Best when the champion list, interview notes, and success playbook should live in one doc with lightweight automations. Like Airtable, it depends on someone feeding it data.

The method works without any of these tools. Add tooling when the evidence outgrows what one person can read.

FAQ: Customer Champion Programs

What is the difference between a customer champion and an NPS promoter?

An NPS promoter gives you a 9 or 10 on a survey. A champion succeeds measurably, expands usage, brings others in, and defends you when things go wrong. Many promoters are satisfied but passive, and research shows a large gap between saying you'll refer and actually referring, so NPS alone should not define your champion list.

How many customers should be champions?

Start with roughly the top 10% of accounts ranked by success signals. If fewer than 5% qualify, your product likely has a success problem. If more than 20% qualify, your criteria are too loose. Small companies can round up slightly, for example 8 of 60, if CS flags proven referrers just below the cutoff.

How do you identify customer champions without new tools?

Score every account on 4–6 signals you already have: feature adoption breadth, month-over-month usage growth, unprompted seat additions, recorded referrals, and low support burden. Weight them out of 100, rank, validate the top 10% with CS, remove false positives, and re-score every quarter.

What are good champion scoring criteria?

A practical starting weighting is adoption breadth 30, usage growth 25, unprompted seat adds 20, referrals or intros on record 15, and low support burden 10. Adjust the weights after one or two quarters based on which signals actually predicted expansion and referrals.

What should a champion program give champions?

Give direct access instead of support tickets, early access to features built from their feedback, real roadmap influence, public and personal recognition, and assets that help them advocate internally, such as case studies, ROI data, and talking points for their boss.

How do you measure a champion program?

Measure network value, not individual LTV. Track direct and indirect referrals, influenced deals, champion expansion rate, champion-sourced revenue, time to champion status, and champion retention compared with non-champions.

Do you need a dedicated champion success manager?

Not at first. The role, managing 15–25 champions instead of 50–100 accounts, makes sense once you have more champions than one person can know well. Before that, one CS lead can own the program part-time.

Sources and Further Reading

Spencer Shulem, Customer-Led Development: How to Build What Your Customers Want When AI Can Build Anything, Chapter 17: Champions
Net promoter score (Wikipedia)
Customer lifetime value (Wikipedia)
Referral marketing (Wikipedia)
Kumar, Petersen and Leone, “How Valuable Is Word of Mouth?” Harvard Business Review, October 2007
Schmitt, Skiera and Van den Bulte, “Referral Programs and Customer Value,” Journal of Marketing, 2011
Keiningham, Cooil, Andreassen and Aksoy, Journal of Marketing, 2007
Review of Accounting Studies (2021), study on executive personal connections and supplier relationships, as cited in Customer-Led Development
BuildBetter

Make churn optional.

Your champions are already telling you who they referred, which teams they rolled out to, and what friction slows them down. It's buried in calls, tickets, and Slack threads. BuildBetter surfaces those signals, ties them to your roadmap, and tells champions when you ship what they asked for.

Make churn optional. Book a demo →