Word of mouth

How to run word-of-mouth like paid media

Pre-set CAC, end-to-end attribution, pay-on-sale: the building blocks that turn word-of-mouth into a measurable channel, with a -26 % CAC.

Virginie Maire
8 min read

Key takeaways

  • In advertising, you discover your CAC after the fact; in recommendations, you set it upfront and never exceed it.
  • On a 10 € CAC set by the brand, 8 € goes to the customer ambassador and 2 € to the platform, with no setup fees.
  • Without a personal link assigned to each share, a recommendation lands in your analytics tagged as direct traffic.
  • A spent paid budget leaves nothing behind; a recommendation budget leaves a first-party database of identified ambassadors.
  • Across our first brands, internal figures: up to 40 % of sales from recommendations and an average -26 % acquisition cost.

Everyone knows word-of-mouth works. Nobody knows how to fit it into an acquisition plan.

That is where the bottleneck lies, and it is technical before being cultural. An acquisition director cannot defend a channel in committee if they cannot report unit costs, volume, or growth trajectory. Faced with a Meta budget that tracks CPA to the decimal, a channel without numbers always loses the trade-off.

Yet the context makes this trade-off increasingly uncomfortable: customer acquisition costs have risen by more than 60 % in five years (Paddle, ex-ProfitWell), and I detailed the reasons behind this ceiling in L’acquisition client est cassée.

So let’s ask the question differently: what if you could manage word-of-mouth campaigns just like paid media? Here are the building blocks required.

How to set your CAC before spending a single euro

You decide it, and the system sticks to it. This is the primary reversal compared to advertising.

In paid channels, you commit a budget and discover your acquisition cost after the fact, subject to competitor bids and seasonality. In recommendations, you set the amount you agree to pay for a new paying customer, and that amount holds.

The reward is configured in three ways:

  • Fixed amount: 5 € per new customer, regardless of the cart size. Clear for everyone, including the referrer.
  • Percentage of the cart: useful when your cart values vary significantly from one product line to another.
  • Tiers: the reward scales with the volume or value generated, to reward consistent ambassadors.

You add restrictions to protect your margin: a payout delay, a minimum order value. The delay covers your return or cancellation window, avoiding paying for an acquisition on an order cancelled three days later.

To choose the number, start from your unit margin and customer lifetime value, not what your competitors pay on Meta. That is the entire point of a channel where you control the dials.

Second direct consequence: you do not pay for exposure. Rewards are only paid out when the objective is met, with no setup fees and no volume commitments. On a 10 € CAC, the breakdown is straightforward: 8 € in rewards goes to the customers who drove the sale, 2 € in commission goes to the platform. Four-fifths of your acquisition budget therefore ends up with real people. Compare that to the share of your Meta budget that ever leaves Meta.

How to track a recommendation end-to-end

With an identifier attached to every share. Without it, steering is impossible, and that is precisely what word-of-mouth has always lacked.

A link sent via WhatsApp, Instagram DM, or Slack lands on your site without a referrer. Your analytics tool classifies it as direct traffic, which accounts for around 37 % of visitors. The details of this black hole are covered in Dark social : pourquoi vos analytics ne voient pas le bouche-à-oreille.

On the customer side, the journey takes five steps:

  1. They visit or purchase on the e-commerce site.
  2. The brand invites them to share to earn a cash reward.
  3. A personal link is automatically generated.
  4. They share via WhatsApp, Instagram, or their preferred channel.
  5. No forms to fill out: a secure app is all they need to track their earnings.

Step three is the critical building block. The personal link makes the chain measurable and tracks it across multiple tiers: when the person who received the link shares it again, the new link stays connected to the original chain. That is what separates an attributed recommendation from a “direct” visit that no one can ever explain.

What does a recommendation dashboard look like?

Like an ads manager, with one extra view. To manage it like paid media, you need the same tools.

Overview: real-time performance tracking. No monthly reporting stitched together after the fact: you see what is happening while it happens.

Campaign creation: objectives, budget, target CAC, restrictions. You launch a campaign on a product, line, or period, exactly as you would build a media plan.

Your ambassadors: a leaderboard of the entire recommendation funnel, with the ability to send push notifications.

This third view has no equivalent in paid ads, and it is the most interesting one. You cannot talk directly to a Meta audience. You can re-engage your top ambassadors, reactivate dormant ones, and notify those who are already selling for you about new arrivals.

On top of that comes the benefit that outlives the campaign. On advertising platforms, most of the value and data remains locked within the ad network. Here, the brand tracks its acquisition performance and owns the data, first-party and in real time.

A spent paid budget leaves nothing behind. A recommendation budget leaves a database of identified ambassadors.

Each campaign therefore enriches an asset you own, reusable across your CRM, segmentation, and subsequent campaigns.

What makes the channel viral, not just measurable

Three design choices separate a clean but flat setup from a self-sustaining channel.

Open to everyone. Anyone who wants to recommend your brand can do so, without vetting or audience thresholds. Traditional referral is limited to existing customers and affiliate marketing to professionals: these two filters are the primary cause of the disappointing activation rates I describe in Pourquoi votre programme de parrainage ne marche pas.

Zero friction. No email, no password, no promo code to dig out of an inbox. Biometrics handles authentication in a single gesture, and it is probably the most underestimated factor in activation rates.

Cash. A cash reward, spendable anywhere and immediately, motivates far more than a store voucher that forces the referrer back to your shop to claim their due. Cardlytics makes the same observation regarding cash back versus brand rewards, and the topic is covered in depth in Cash, points ou bons d’achat.

CriterionPaid adsTraditional referralManaged recommendations
Who can participateNo one, you buy ad spaceYour existing customersAnyone who wants to recommend
Cost knownAfter the campaignVariable, often store creditSet upfront, by the brand
Payment triggerImpression or clickReferral signupConfirmed sale
AttributionModeled by the ad networkPromo code, often lostPersonal link, multi-tier
DataOwned by the platformPartialFirst-party, owned by the brand

The combination of these three choices produces the network effect: brands join the platform, customers become ambassadors, recommendations generate sales, and each sale creates new ambassadors.

What results to expect, and how fast?

The academic precedent is encouraging. Across a German bank’s referral program tracked over six years, referred customers cost 20 € less to acquire than others, and the program generated a 60 % return on investment despite a 25 € reward paid to each referrer (Schmitt, Skiera & Van den Bulte, Journal of Marketing).

On our side, across our early brands, the setup produces up to 40 % of sales from recommendations and an average 26 % reduction in acquisition costs. These are our internal, unaudited figures: attributed sales and acquisition costs observed in the back office of the relevant brands, to be taken as estimates. It is the kind of data you can bring to an executive board meeting without spending ten minutes explaining the methodology.

Two caveats, because this channel is not magic. It depends on the quality of your product: an unhappy customer base will not recommend anything, regardless of the reward amount. And it ramps up gradually, following the pace of your customer base, whereas a paid campaign delivers volume from day one. For a cold launch, paid ads remain the fastest lever. The right economic perspective on this is in Le bouche-à-oreille est gratuit : pourquoi payer ?.

What is new here is not word-of-mouth: it is the infrastructure. What was missing was the tool to track it, reward it, and scale it reliably. The next acquisition channel is recommendations.

Frequently asked questions

Can you really manage word-of-mouth like a paid campaign?

Yes, provided three conditions are met: an acquisition cost set upfront by the brand, a personal link that attributes each sale to the person who triggered it, and a real-time dashboard to make decisions. You then follow the same workflow as in paid ads: launch, measure, adjust, stop.

How do you set the amount of a recommendation reward?

Based on your unit margin and customer lifetime value, not market bids. The amount can be fixed, set as a percentage of the cart, or structured in tiers. On a 10 € acquisition cost defined by the brand, 8 € goes to the customer ambassador and 2 € to the platform.

What is the difference between a referral program and a recommendation channel?

Referral is restricted to existing customers, often rewards in vouchers, and relies on a promo code that the referee must remember to enter. A recommendation channel is open to anyone willing to participate, rewards in cash after the confirmed sale, and attributes each share via a personal link traceable across multiple tiers.

Word-of-mouth is free, why pay for tooling?

Because spontaneous word-of-mouth is neither measurable nor manageable: it arrives as direct traffic, with no identifier, and you can neither scale nor forecast it. What you pay for is attribution, rewarding the person who drove the sale, and campaign management, on a 100 % performance-based model.


I am Virginie Maire, co-founder of Frak Labs, turning your customers into a scalable, profitable, and authentic acquisition channel. A committed entrepreneur and mother of two, I have navigated media, social networks, influence, and e-commerce for 20 years… and I still love every minute of it!

TopicsCACcustomer acquisitionattributionfirst-party dataadvocacyword-of-mouth

Try Frak, the Recommendation Economy platform

Zero fixed costs. One 20% commission on performance. Cashback for buyers, cash for recommenders.

Download the app