Can word-of-mouth marketing really be quantified?

Partly. It cannot be as precise as advertising. An ad path is short (click, then convert); a word-of-mouth path is long: a consumer sees a discussion, searches the brand weeks later, browses social media in between, and may finally order through another channel. Conversions you can attribute directly to a word-of-mouth link or discount code will always be understated, so read them together with indirect signals such as branded search volume, review-keyword rankings, and a "how did you hear about us" survey, rather than relying on any single number.

Which metrics matter most in word-of-mouth marketing?

Look at four groups, depending on your stage: discussion volume and brand mentions (are people talking about you), branded search volume (are people actively looking for you), referral traffic and conversions (are people visiting and buying, tracked with UTM parameters, dedicated discount codes and GA4 conversions), and cost per order (total spend divided by attributable orders). Which group deserves the most attention depends on whether you are building awareness, building trust, or pushing for sales.

How do UTM parameters and dedicated discount codes work in word-of-mouth marketing?

A UTM parameter is a tag added to a URL so GA4 knows which platform and campaign a visit came from. A dedicated discount code is a promo code used for one campaign or one partner, which catches orders from people who read the content but did not click the link. Two cautions: many forums restrict external links, so check the rules before posting; and a discount code is an offer open to everyone who sees the content and must never be conditional on writing a positive review. Neither method captures every conversion, so treat both as a trackable floor.

How do I calculate the return on word-of-mouth spending?

The most practical measure is cost per order: divide the total spend on a campaign (service fees, product samples, shipping, content licensing, internal staff time) by the orders attributable to it in the period, then compare with ads or other channels. The denominator only includes trackable orders, so it understates word-of-mouth contribution. Report branded search changes and survey results alongside it, and state the attribution limits.

I set up tracking and still cannot tell where many orders come from. Is that normal?

Yes. Word of mouth often makes people remember a brand and act later, on another device or channel, and analytics tools usually log those conversions as direct traffic, organic search or ads. The answer is not to force every order into an attribution bucket. Report a trackable floor plus supporting indirect signals, and use a checkout survey to capture self-reported sources.

How to Measure Word-of-Mouth Marketing ROI

Plenty of brands run word-of-mouth marketing for six months, and when the boss asks how it went, the best the marketer can offer is “there’s more discussion out there when you search, so it feels like it’s working.” That answer will not convince anyone to keep funding the program.

Word of mouth is harder to quantify than advertising, but not impossible. It needs a different approach. This guide covers four things: which metrics to watch, how to set up tracking, how to calculate cost, and where attribution runs out, so you can build a performance report you can actually defend.

Start by understanding how word-of-mouth attribution differs from ad attribution

An ad path is short: click the ad, land on the site, buy. That conversion traces back to the individual ad.

A word-of-mouth path usually looks like this: a consumer reads a genuine user review on a forum or social platform, does nothing at the time, searches for the brand weeks later, lands on the site from that search, leaves after a brief look, and eventually clicks a retargeting ad and buys.

In that path the final conversion is credited to “advertising,” even though word of mouth is what started the purchase intent. That is why measuring only direct conversions undervalues word of mouth. The honest flip side is that nobody can say exactly what share of an order came from word of mouth. You can only approach the answer with several signals read together.

Metric definition table: how each metric is calculated and where the data comes from

Define the terms first, so reports do not mean different things to different people. The table below lists six common metrics and the brand stage each suits.

Metric	How it is calculated	Data source	Suitable stage
Exposure	Times the word-of-mouth content was seen (post views, search impressions)	Platform dashboards, Google Search Console	Just starting, need to be seen
Discussion volume	Posts and comments mentioning the brand or product in a period, split into positive, neutral and negative	Social listening tools, manual search	Building awareness and trust
Branded search volume	Search impressions and clicks for the brand name (Chinese and English spellings, abbreviations, common typos), viewed monthly	Google Search Console	Awareness-building; shows whether people now look for you
Referral traffic	Sessions arriving on your site from word-of-mouth content	GA4 with UTM parameters	You have content and links are allowed
Conversion	Purchases, form submissions or bookings from that traffic; discount code redemptions	GA4 conversion events, discount code report in your order system	Pushing for sales
Cost per order	Total campaign spend divided by attributable orders in the period	Finance and order systems	Reporting upward, comparing channels

A few definitional reminders:

Discussion volume is not performance. More discussion with a rising negative share is a warning, not progress, so always pair it with sentiment classification.
Lock branded search to a fixed keyword list and compare using the same list each month, or the numbers are not comparable.
Count the full numerator in cost per order: service fees, product samples, shipping, content licensing, and internal time for reviewing and replying. Counting only the service fee makes word of mouth look cheaper than it is.

Tracking setup, step by step

Step 1: Agree a UTM naming convention

UTM parameters are tags on a URL that let GA4 tell traffic sources apart. The common failure is not technical but inconsistent naming: one person writes ptt, another PTT, another forum, and reports fragment into pieces. Write the convention down first:

Parameter	Purpose	Naming guidance	Example
utm_source	Source platform	Lowercase, one fixed name per platform	`dcard`, `instagram`, `blog`
utm_medium	Content type	A small fixed set of values	`koc`, `forum`, `kol`
utm_campaign	Campaign period	Year and quarter plus short name	`2026q4_newlaunch`
utm_content	Tells apart creatives within a campaign (optional)	Creator code or asset number	`creator07`

Example URL:

https://example.com/product?utm_source=dcard&utm_medium=koc&utm_campaign=2026q4_newlaunch&utm_content=creator07

Notes:

Check platform rules first. Some forums restrict or ban external links and commercial collaboration posts. Do not break the rules for the sake of tracking; where links are not allowed, rely on discount codes and surveys.
One utm_content per partner, so you can later see which partners bring better-quality traffic.
Keep a shared spreadsheet of every URL generated, to avoid duplicates and typos.

Step 2: Set up dedicated discount codes

UTM only catches people who clicked the link, but many people read the content and then search the brand or type the address themselves. A dedicated discount code covers part of that gap:

One code per campaign or per partner, for example DCARD2610, with redemptions and amounts reported by your order system.
Make the code available to everyone who sees the content, and state the terms and expiry in the content.
A discount code must never be a reward for writing a positive review. It is an offer to consumers, not payment for review content. If you give product samples or payment to creators, disclose the relationship in line with Taiwan endorsement guidelines (see our guide to word-of-mouth marketing compliance in Taiwan).
Codes that get reposted to coupon sites distort the data, so set expiry dates and usage caps, and check periodically for unusual concentration.

Step 3: Set up conversion events in GA4

Decide which actions count as conversions: purchase, inquiry form submission, booking, phone or LINE button click.
Mark those events as key events (conversions) in GA4 and confirm they fire before trusting the reports.
In the traffic acquisition report, break results out by utm_source, utm_medium and utm_campaign, and compare sessions, engagement and conversions per channel.
For e-commerce, pass order value back to GA4 so you can see revenue per channel.
After setup, walk through the flow yourself with a test link to confirm the parameters are recorded before going live.

Step 4: Watch branded search volume

Verify your site in Google Search Console (free).
Build a fixed branded keyword list: Chinese and English names, abbreviations, common typos, and combinations such as “brand + reviews” and “brand + recommendations.”
Record a baseline before the campaign (the month before launch), then check again during the campaign and one to two months after, and compare the direction of change.
Log concurrent ads and promotions. If heavy advertising ran at the same time, a rise in branded search cannot be credited entirely to word of mouth, and the report should say so.

Step 5: Add a “how did you hear about us” survey

Add an optional question at checkout: “How did you hear about us?” with options such as forum or social post, friend recommendation, search, ad, other. This captures what analytics cannot see, especially word-of-mouth influence that carried no link and no code. Self-reported sources are subject to memory bias, so treat them as corroboration rather than precise figures.

Discussion volume and sentiment need ongoing records. Choose by scale:

Approach	Suited to	Notes
Manual search plus a spreadsheet	Brands on a tight budget or just starting	Use the same keyword list monthly and record new posts and sentiment by platform
Free or entry-level listening tools	Small and mid-sized brands wanting to save labor	Features, platform coverage and pricing vary and change; verify before you commit
Paid monitoring services	Brands needing cross-platform, long-term, detailed reports	Usually monthly or annual fees; confirm the platforms your audience uses are covered

For tool selection in more detail, see our guide to brand monitoring tools. Whatever the tool, the core is recording monthly with the same definitions; the trend matters more than any single month.

Illustrative calculation: working out cost per order

All figures below are illustrative assumptions used to show the method. They do not represent any brand’s actual results. For estimating the cost structure itself, see the complete guide to word-of-mouth marketing costs.

Illustrative assumptions: a brand runs a three-month word-of-mouth campaign for a new product priced at NT$1,000. Margin is not discussed here.

Item	Illustrative amount
Service fee (recruitment, review, reporting)	NT$60,000
Product samples plus shipping	NT$20,000
Content licensing and other	NT$5,000
Internal staff time (about 30 hours, illustratively valued at NT$500 per hour)	NT$15,000
Total spend	NT$100,000

Suppose the trackable results over three months are (again illustrative):

Tracking method	Attributable orders
Orders after UTM-tracked visits	40
Dedicated discount code redemptions	25
Total (assuming no overlap)	65

Cost per order = 100,000 / 65 = about NT$1,538.

Two problems must be stated alongside that number:

It is a ceiling, not the true value. The denominator only counts trackable orders. People who read the content and searched the brand a few days later were not counted, so the real cost per order is usually lower than NT$1,538, though we cannot know by how much.
Cost per order alone is not enough. With a NT$1,000 product and NT$1,538 marketing cost per order, a first purchase alone clearly does not pay back. Whether that is acceptable depends on repeat purchase, order value and the long-term value of the content, all of which must come from your own data rather than borrowed ratios.

The recommended presentation: state the trackable cost per order first, then place branded search change, discussion and sentiment trends, and survey results alongside it, so decision-makers see a set of signals rather than one over-interpreted number. If you are comparing approaches (for example KOL versus KOC), both must use the same tracking method and the same observation window.

Attribution limits: say them plainly

Before reporting to your boss, put these limits up front. It increases credibility:

Cross-device and cross-time behavior is invisible. Seeing on a phone and ordering on a laptop often appears as two people.
Direct and organic search often hide word of mouth’s contribution. Word of mouth makes people remember the brand and return later, and tools file that as “direct” or “search.”
Platform link restrictions. Where links are not allowed, UTM fails and only discount codes and branded search remain.
Overlapping campaigns cannot be separated. When ads, promotions and KOL posts run together, the rise in branded search is a shared result and cannot be split precisely.
Discount codes leak. Once leaked, redemptions no longer reflect the original content’s contribution.
Self-reported sources are biased. People remember the most recent and vivid touchpoint.
Negative word of mouth is measured too. Rising discussion is not automatically good; check sentiment and topics.

The working principle: treat trackable numbers as a floor, treat indirect signals as direction, present them side by side, and draw no conclusion the data cannot support. If a vendor promises that “every order can be precisely attributed to word of mouth” or “guarantees search volume will grow by a set amount,” ask for evidence.

A standard format for the monthly word-of-mouth report

A monthly report should cover:

Metric	This month	Last month	Change	Source and notes
Branded keyword search impressions	—	—	—	Google Search Console, fixed keyword list
Review-keyword rankings	—	—	—	Monthly manual record of positive and negative share on the first pages
Discussion volume (positive / neutral / negative)	—	—	—	Listening tool or manual log
Sessions from word-of-mouth links	—	—	—	GA4 + UTM; note platforms where links are not allowed
Discount code redemptions	—	—	—	Order system
Attributable orders and cost per order	—	—	—	Mark as trackable floor
Survey: self-reported source share	—	—	—	Optional checkout question; corroboration only
Negative reviews and handling status	—	—	—	Including those responded to and resolved

Quantifying word-of-mouth ROI is not about one number. It is a set of signals: the direction of branded search, the quality of content on review keywords, trackable traffic and orders, customers’ self-reported sources, and a total spend you can account for. Present them together, with the limits stated honestly, and you have a performance report that holds up.

If you want a complete word-of-mouth measurement framework, or need to turn your existing data into a report you can take to management, talk to a NETVANA consultant. For how to split budget between word of mouth and ads, see Word of Mouth vs. Paid Advertising: A Budget Allocation Guide.

Further reading: before you hire a word-of-mouth marketing agency, make sure you pick one that can actually deliver performance reporting. How to Choose a Word-of-Mouth Marketing Company: Avoiding Content Farms and Black Boxes lists the questions you need to ask. For deciding where to spend once you can measure, see Word-of-Mouth Budget Planning for Small Businesses. For a leading indicator you can track over years, see What Net Promoter Score Actually Measures; and for social metrics that map to business outcomes, see How to Set Social Media KPIs. For what you can and cannot measure about AI visibility, see Brand Visibility in the Age of AI Search. For the metrics a referral program should track, see Designing a Referral Program. For the touchpoints a customer passes between first hearing about you and ordering, see Word-of-Mouth Touchpoints in the Customer Journey.