Insight / Featuring Rand Fishkin

What Rand Fishkin and Amanda Natividad would question in a customer survey

In Zero Click Marketing, Amanda Natividad and Rand Fishkin examine what goes wrong when a customer’s recollection becomes the explanation for a sale.

Social Signal Playbook EditorialAnalysis · AI-assisted · Nov 8, 2026 · 3 min read

The Thesis

A customer can truthfully remember a podcast and still be unable to tell you what caused a purchase. In the October 13 episode of Zero Click Marketing, host Amanda Natividad and her guest and coauthor Rand Fishkin examine the cost of asking buyers to assign credit. They also address a tempting fix. Letting customers type their own answers changes the form, but the customer still has to reconstruct what happened.

Context & Analysis

Decide what you need a customer to remember before asking. A buying occasion, a remembered source and a recorded visit answer different questions.

Where to find the argument

Natividad’s episode guide places their discussion of the familiar how-did-you-hear-about-us question at 14:24. They turn to the limits of an open text box at 20:12, discuss an alternative question at 25:28, and consider what a poor question costs at 30:09. A survey takes some of a buyer’s time. Using that space to ask for a channel name also means passing up a chance to learn which problem brought the person to the shop.

Why a familiar name can win

Fishkin gave a concrete example in a separate November 2 conversation with Ross Hudgens. According to Siege Media’s notes, he described an experiment in which 30% of a brand’s customers credited a well-known food personality who had never mentioned the brand. The publisher places that story at 02:26. This is Fishkin’s account of one experiment. It illustrates how plausible an answer can sound even when the supposed referral did not occur. It supplies no estimate of the error rate for another brand.

Ask about a moment the buyer can describe

Suppose a music teacher is trying to understand why adults sign up for a first lesson. One question we would try is “What was happening when you decided to look for lessons?” Someone might describe inheriting a piano, hearing a friend play or finding time after a shift change. Each answer suggests a different conversation the teacher can have with a prospective student. This is our example question. Listeners can hear the coauthors’ own discussion at 25:28.

What to do with a messy answer

Suppose the student writes that a friend sent them a performance, they watched several more over the summer, and they finally booked when the teacher offered evenings. A single box marked “friend” would discard the scheduling detail. In this example the teacher could keep the original wording, record the friend as a remembered introduction and ask whether the evening schedule made the difference. The answer might change which hours the teacher offers. No attribution model is needed to act on that piece of customer research.

Keep the evidence attached to its question

The teacher can record the student’s story beside the link on the booking form. A story about inheriting an instrument describes an occasion. A tagged search visit identifies the route into the form. A mention of a friend describes something the student recalls. Keeping those observations separate makes it possible to improve the lesson description without awarding an entire booking to whichever name was easiest to remember.

What Has Changed Since

On November 2, Fishkin returned to the survey question in a separately published interview with Ross Hudgens. Siege Media’s chapter list locates the food-personality experiment at 02:26 and a discussion of incrementality tests at 06:58. That later source gives a concrete case for the concern raised with Natividad on October 13.

Frequently Asked Questions

Who is speaking in the October episode?

Amanda Natividad hosts Zero Click Marketing. Rand Fishkin returns as her guest and coauthor to discuss their book, customer research and how he is learning to market a game.

Does the 30% example come from the October episode?

The cited example comes from Siege Media’s description of Fishkin’s separate November 2 interview with Ross Hudgens. The October episode’s own chapter guide covers the limits of post-purchase surveys and free-text replies.

Where do the hosts discuss open-text answers?

The October chapter guide places that discussion at 20:12. A person can name a source outside a preset list and still have an incomplete memory of which encounters influenced the decision.

How can a team use survey answers when it also has web analytics?

Keep the buyer’s account as testimony and the visit path as an observed event. If the purchase followed a search visit, a buyer’s account of an earlier recommendation can explain why they searched. Either observation on its own leaves part of the sequence unknown.

Works Cited & Evidence

A source attribution identifies where a claim came from. It does not independently verify the claim. Inspect the original context and limitations. Editorial methodology · Corrections

The Hardest Parts of Zero Click Marketing (Rand Fishkin returns!)

primary source·Tier 3: Low-Authority Context·Zero Click Marketing·Oct 13, 2026

Amanda Natividad names Fishkin as her coauthor and guest. The published chapter guide places attribution, free-text answers, an alternative question and opportunity cost at 14:24, 20:12, 25:28 and 30:09.

View Source

Rand Fishkin: “How Did You Hear About Us?” Is Lying to You

supporting source·Tier 3: Low-Authority Context·Content and Conversation: Organic Growth Insights from Siege Media·Nov 2, 2026

In a separate November interview hosted by Ross Hudgens, Siege Media places Fishkin’s account of the food-personality survey at 02:26.