Most DTC brands that come to us are not struggling because their product is bad. They are struggling because the creative brief was written from the brand's perspective — what the team thought customers valued — while the customer was somewhere else entirely, writing in the wild what they actually wanted.
The brand in this case study had strong reviews, real retention, and CPMs creeping past $20. The product was fine. The creative was the problem. We ran the methodology described in our VOC ad copy framework on a real engagement, and the result was 236 categorized comments → 10 ranked hooks → one full creative brief the media buyer could ship against the same week.
The Extraction in Numbers
The dataset was not built from surveys or focus groups. It was scraped from the surfaces where the brand's buyers were already talking without being asked.
- 236+ comments collected in total.
- 3 platforms — the same three surfaces the rank-and-rent framework targets.
- 112 from TikTok comment threads, App Store reviews, Reddit threads, and BGG community posts — the place where superfans and detractors speak in full sentences.
- 8 verified Amazon reviews (4.9-star product, so noise was minimal) — every line is from someone who paid.
- 116 from Meta and Facebook — full reviews, product page comments, Reel reactions, and Ads Library entries on competitor ads.
That split matters. It is not a vanity total. Every comment was categorized before counting.
Methodology, Step by Step
The full framework lives in the post linked above. Here is how it looked when applied to a real engagement with a real brief, real timeline pressure, and real money on the line.
1. Scrape across three surfaces — not one
Amazon reviews are the richest single source because verified-purchase customers are solving a social problem: explaining to a stranger whether the product is worth it. That framing produces uniquely useful copy — what they tried before, what surprised them, which features mattered and which didn't. Three-star reviews are often the most useful of all because they describe the gap between what was expected and what was delivered.
TikTok comments capture a different demographic and a different voice — faster, more casual, more emotionally raw. The comments that start with "wait" or "ok but" tend to signal genuine surprise or objection. Comments that tag friends reveal social proof dynamics — what makes someone want to share. We pulled from organic videos about the product and the top three competitors, not just our own brand.
Meta and Facebook ad comments are the highest-signal source for objections. People who comment on ads do so because something the ad said or showed provoked a response — agreement, skepticism, a question, a claim. Competitor ad comments via the Facebook Ad Library are particularly valuable because they surface the objections standing between your category and a purchase decision.
2. Categorize every line into four buckets
Raw VOC is not ad copy. It is input. Each line from the 236 was sorted into one of four categories: hooks (language that stops thumbs), objections (reasons not to buy), desire statements (what customers want but don't yet have), and competitive positioning (how customers describe alternatives, including competitors by name). Everything went into a bucket. Nothing got skipped because the line was short or ungrammatical — short, ungrammatical lines are often the most candid.
3. Rank by cross-platform frequency and emotional specificity
Lines that appeared across more than one surface got weighted higher. A phrase that showed up on TikTok and in Amazon reviews is more reliable than one that showed up in only one place. Within each surface, ranking was by frequency plus emotional specificity — generic pain underperforms specific pain, and a number beats a feeling when the audience has seen a thousand ads.
4. Map the top lines to ad formats
The final step was format-by-format deployment. The strongest TikTok native line went to a TikTok spark ad. The strongest objection-handler went to a Meta retargeting carousel. The strongest desire statement went to a Reel script. The strongest competitor contrast went to a conquest ad in the same audience the competitor was buying. Every hook in the brief had an assigned format so the media buyer did not have to make that decision again.
Anonymized Quote Samples
Every quote below is a real line from the engagement, lightly anonymized. We are not inventing customer language for marketing copy. That would defeat the entire point of VOC — the language is the asset.
"My son bought it. I've been stealing it from him to play in the mornings."
Source: App Store review, echoed in a Facebook product-page comment. Attributed to a 50-something buyer who became the primary user of a product she did not buy for herself. This line did more cold-audience work than anything the brand's in-house copywriter produced.
"We bought it for family game night and ended up playing it on date night too. Then I took it on a work trip and three coworkers bought it the next week."
Source: Amazon review (4-star). Note the three-step adoption arc inside a single review — initial context, secondary context, third-party conversion. This is the kind of data a survey never produces because customers do not volunteer it when asked.
"Honestly I was going to cancel the order after reading the price. Then I saw the reviews and figured I'd try it once. I've bought two more as gifts since."
Source: Verified Amazon review. The objection ("too expensive") is named and dismissed by the same customer — exactly the structure a good retargeting ad copies. No copywriter invented this frame. It is sitting in the review section.
What the Brief Produced
The output was not a slide deck of observations. It was a creative brief the in-house team could execute against the same week. Seven sections, each grounded in verbatim customer language and ready to deploy.
- 10 ranked hooks — ordered by cross-platform emotional punch, with notes on which ad format each works in. Not a brainstorm. A ranked list produced by frequency and specificity analysis.
- 5 ad angles — distinct positioning themes (Morning Ritual, Gift Magnet, Screen-Free Antidote, Brain Sharpener, Founder Story), each backed by three cross-platform customer quotes.
- 5 objection handlers — top friction points paired with reframe language ready for ad copy and landing-page FAQ. One of them was "out of stock" — reframed into a conversion event, not a lost sale.
- 7 desire statements — what paying customers want but do not have yet, each mapped to a product or positioning opportunity. One produced an immediate SKU recommendation.
- 5 audience personas — built from actual buyer language, not demographic assumptions. Each with a platform strategy, ad format, and lifetime-value estimate.
- 1 competitor positioning map — how customers compare the product to alternatives in their own words, with four ready-to-use competitive copy lines.
- A format-to-hook deployment table — every ranked hook mapped to TikTok spark, Meta feed, Instagram Reel, Amazon sponsored, or retargeting creative, so the media buyer did not need a second brief to start running it.
Read the Case Study, Run It on Your Customer Data
If you want the full methodology before your next creative sprint, the long-form case study breaks down the challenge, the process, and the results with the actual quote pull-quotes in context:
For a condensed results-only view focused on the conversion output:
→ Want this done on your customer data?
Related reading: How to Mine Customer Reviews for Ad Copy That Converts — the framework this case study was built on.