Resources

Tools to build your referral program.

Calculators, playbooks, and AI skills, free for operators shipping referral programs.

Tools

Coming soon
Best-in-Class Referral UX

40 B2B referral programs mapped and annotated

By Kate Syuma. Incentive types, UX mechanics, reward structures: all in one Miro board.

One email when it's ready. Nothing else.

Claude Partner Skill

Partner Radar: source and score partners with AI

A ready-to-install Claude skill: a go/no-go read on partner-led growth, scored partner segments, 25 named leads with contact routes and hooks, outreach copy and a launch plan.

Cello (Powerplay GmbH) stores your email for this download. Privacy policy.

New tools as we build them

We're adding calculators and playbooks to this page. Tell us where to send them.

Cello (Powerplay GmbH) will email you when a new tool goes live on this page, and store your address as a record of this request. Cello keeps your address while this list is active and deletes it whenever you ask. Every email has an unsubscribe link. Privacy policy.

Reference

Lookups pulled out of the chapters, ordered by chapter. Some open the chapter they came from, others open a full reference table further down this page.

Benchmarks

B2B Referral Program KPI Benchmarks

Referenced from Chapter 01: the KPIs worth tracking for each of the three referral program types. Fuller benchmark detail lives in Cello's B2B referral metrics guide.

Program typeKPIs to measure
User Referral ProgramsActivation rate, sharing rate, number of new users, purchase rate from user referrals, total user referral ARR.
Affiliate Referral ProgramsClick-through rate (CTR), conversion rate, affiliate earnings per click (EPC), total affiliate referral ARR.
Influencer Referral ProgramsEngagement rate on sponsored content, conversion rates from influencer referrals, brand awareness metrics (# of impressions, # of shares, etc.), total influencer referral ARR.
Unit economics

Referral Program Unit Economics

Referenced from Chapter 01: the numbers to set your reward against before you commit to a payout.

InputWhere it should land
Target LTV/CAC3:1 or higher. A referral program that cannot clear this is paying too much per referral or converting too few of them. Note that referral CAC is usually counted as rewards only: platform fees, engineering time, program management and the margin given away in the referee discount are rarely in the number.
Reward capStart from ACV and work backwards to the headline number users actually see, for example a percentage of MRR up to a limit such as $10,000.
Reward basisTie rewards to generated revenue wherever possible. It carries built-in fraud protection and removes the payback period entirely.
Sharing rate5 to 15% of active users sharing a link is the healthy band.
Invites per referrer2 to 5 invites on average from each user who shares. This is a mean over a heavily skewed distribution: a small group of advocates sends most of the invites, so plan against the shape rather than the average.
Invite to paid conversion25% and above for free-trial products. Referred free users converting to paid at over 30% is a free-trial number; against a freemium base it would be extraordinary. In Cello platform data (n = 4 million B2B SaaS referral users) referred traffic converts up to 5x better than paid and high-intent SEO in the same companies. Growth Unhinged publishes a separate 5x, comparing card-required trials with trials without a card: different comparison, different claim.
Referred user qualityWharton research: referred users churn 18% less and carry 16% higher lifetime value. They also carry a contribution-margin advantage early on, but it decays and is gone by roughly month 29, so do not model it as permanent.
Glossary

Referral Metrics and the AUCC Benchmark

Referenced from Chapter 03. Everything you need with a spreadsheet open: what each funnel metric means, what best in class looks like across Cello platform data covering 4 million B2B SaaS referral users, and the quantity and quality benchmarks worth tracking alongside them.

The funnel metrics
MetricDefinition
Sharing ratePercentage of users who share a link with potential new users.
Unique clicks per shared linkNumber of potential new users who click the referral link.
Sign-up ratePercentage of potential new users who click through to sign up for the referred tool.
Purchase ratePercentage of new users who convert to a paid plan.
Annualized user to customer conversion (AUCC)New paying customers generated via referrals per 1,000 existing users over a year, expressed as a percentage of that user base. AUCC is Cello's own term rather than an industry standard: searching it returns only cello.so. Best in class is 49%, meaning paid users grow by roughly half again over a year through referrals alone. Companies such as tl;dv, Blockpit, Typeform and Fellow see 10 to 15% additional monthly revenue growth from referrals.
Quantity and quality benchmarks

The funnel metrics tell you what happened. These three tell you whether the program is attracting the right users in the right volume.

BenchmarkWhat it tells you
Participation ratePercentage of active users who have shared at least one referral. This is the read on how well you recruit referrers out of the user base you already have.
Invites sent per userAverage number of invites each referrer sends. User referrals are a brand-building channel as much as an acquisition one: most people have a following and share one-to-many across WhatsApp, Slack and LinkedIn, so reach is part of the return.
Visit to signup, signup to activationConversion at each step after the invite lands. Referred users often need more activation energy and know less about the product than users from other channels, so activation is the step that usually leaks.
Tool

Calculate your word-of-mouth coefficient

Two numbers a month in, one number out. Nothing is submitted anywhere and nothing is saved: it recomputes as you type.

01

Your numbers

Two numbers per month: active users, and new users arriving through direct traffic plus brand search. Where those come from is in section four. Everything recomputes as you type.

Monthly active users, new word-of-mouth users, and the resulting per-month coefficient. This table is also the data behind the validation chart.
MonthActive usersNew WoM usersCoefficientRemove
0.042
0.044
0.041
0.043
0.042
0.042
02

Your coefficient

Word-of-mouth coefficient
0.042

Volume-weighted: total word-of-mouth users (1,377) divided by total active users (32,500) across all 6 valid months, not the average of the per-month ratios. Bigger months count for more.

Per 1,000 active users
42
new WoM users per month
Annualized
50.8%
of the active base, per year
0.99
active users predict WoM intake
The relationship holds.

An R² of 0.99 means active users explain most of the month-to-month movement in your word-of-mouth intake, so the coefficient is stable enough to plan against. Annualized, word of mouth is worth 50.8% of your active base per year, which is at or above the Atlas's best-in-class AUCC benchmark of 49%. Treat that as a direction, not a like-for-like score: AUCC counts customers, and this counts users.

03

Step 4: validate it

Each dot is one month. If active users really drive your word-of-mouth intake, the dots sit close to the line.

↑ New WoM users
0831662483313,8104,6555,5006,3457,190line of best fitJanuary: 4,200 active, 176 WoM, coefficient 0.042February: 4,650 active, 203 WoM, coefficient 0.044March: 5,100 active, 211 WoM, coefficient 0.041April: 5,600 active, 239 WoM, coefficient 0.043May: 6,150 active, 260 WoM, coefficient 0.042June: 6,800 active, 288 WoM, coefficient 0.042June
Active users →
04

Where the two numbers come from

Reforge's method, run light: Google Analytics and nothing else. It assumes active users predict new word-of-mouth users. Start here, and only tighten the definitions if the relationship does not hold.

StepWhat to do
1. Define the denominatorSegment returning users in Google Analytics. That is your active-user count for the month.
2. Define the numeratorIdentify new users arriving through direct traffic and brand search. That is your word-of-mouth intake.
3. Run the analysisExport both series by month and put them in the table above. The coefficient is the division.
4. Validate itCheck R² on the plot. At or above 0.7 the relationship holds and you can forecast against it.

If the light version is not predictive, tighten the definition of an active user around the actions that plausibly generate a recommendation, and run it again. Only mature businesses with genuinely complex marketing mixes need the heavy, multivariate version. Start simple and scale the method as the number starts mattering to your planning.

Full reference

Affiliate budget: rates, ranges and worked examples

The canonical home for the affiliate budget lookup tables. Chapter 08 keeps the method and the two inputs that set the order of magnitude; the full set lives here.

Commission forms and rates
FormHow it pays
Pay-per-clickThe affiliate gets a reward for every successful unique click on a referral link.
Pay-per-leadThe affiliates get a reward for an inbound lead or a specific event.
Pay-per-saleThe affiliates get a percentage of the revenue generated by new paying customers.
Typical rangeSaaS pays the highest commission fees of any category. The rate Cello sees in practice is 15% to 30%. The higher your CLV, the higher the absolute payout can be and the lower the percentage needs to be.
Budget inputs
InputWhat to work out
Average Customer Lifetime Value (CLV)Understand how much revenue the average customer brings in over their lifetime. This is the ceiling every other number sits under.
Profit marginsConsider your profit margins on the products or services being promoted. You'll want to ensure that you're still maintaining a healthy profit after paying out commissions.
Market competitionResearch what your competitors offer in their affiliate programs. To attract top affiliates, your offer needs to be competitive.
Program management costsAccount for the cost of running the program: software, personnel, and marketing materials.
Estimated acquisition through affiliatesEstimate how many new customers you expect to acquire through the affiliate program.
Trial and errorStart with a budget you're comfortable with and adjust as you go. Set up two campaigns and A/B test them to understand what actually incentivizes your user base.
Structural variables
VariableWhy it changes the payout
Two-sided rewardsIncentivizing the referee with a discount as well as the affiliate makes it more attractive for both parties: the affiliate participates, and the new user has a reason to sign up.
Retention-based pay-per-saleOffer revenue shares that extend over time, with an initial and possibly reduced ongoing payment, for example 20% of monthly generated revenue. Recurring fits the subscription model and doubles as a fraud deterrent, since a fraudulent deal that churns after a month stops paying.
Payout thresholdsEstablish a minimum earning for withdrawal to motivate continuous content creation, like Fiverr's $100 threshold, which is a consumer-marketplace bounty rather than a SaaS commission benchmark.
Gamification and tieringIntroduce reward levels for hitting milestones, and add leaderboards and performance boosters to sustain engagement.
Full reference

The full affiliate type comparison

The canonical home for the affiliate type taxonomy. Chapter 08 keeps the six types and what each one is; every attribute lives here.

Affiliate content creators
AttributeNewsletter publisherBlogger collaborationPodcast hostsSocial media creators
WhatPlacement of an affiliate link in a newsletter (e.g. newsletter sponsorship)Placement of an affiliate link in a blog post (e.g. Top 10 AI tools)Short ad placements plus an affiliate link in the podcast bioLink placements in posts (LinkedIn, X, YouTube)
ProTargeted reach plus high conversion ratesHigh trust and authority within blogger networksBrand awareness and high trust in the host's communityBroad audience and potential reach
ConShort time of visibility plus costs not linked to channel performanceTime-consuming to create content and to be mentionedSlow-burn approach plus costs not linked to channel performanceLess control of content plus costs not linked to channel performance
ReachMediumLow to mediumLow to mediumHigh
CostsNiche/vertical products = more targeted affiliates (higher cost). Broad/horizontal products = less targeted affiliates (lower cost).Same as newsletter publisherSame as newsletter publisherSame as newsletter publisher
Effort (own content creation plus affiliate maintenance)MediumHighMediumLow
ExampleLenny's NewsletterNeil PatelSaaStr podcastMarina Mogilko
Suited for: for each of these types of content creators you will find affiliates for either niche/vertical or broader/horizontal products. Keep in mind: the audience of the content creator must match your ICP.
Affiliate hubs
AttributeCoupon and deal websitesAffiliate marketplaces
WhatPlace affiliate links on coupon websitesAccess a wide range of affiliates
ProAttracting price-sensitive customers looking for a deal, plus high conversion ratesBroad access to various affiliates plus acquisition cost control
ConMay devalue the product's perceived value over timeNetwork fees can reduce overall margin
Suited forEarly-stage SaaS with a horizontal targeting trying to acquire their first customersPost-PMF SaaS looking for a diverse range of affiliate types, those new to affiliates
ReachMediumMedium
CostsLowLow
Effort (own content creation plus affiliate maintenance)LowMedium
ExampleAppSumoimpact.com