What I Found in 320 Comments on My Crest Campaign
The comments on my Crest campaign revealed shopping intent without proving completed sales. Creator measurement starts with the job assigned to the video.
My Crest campaign for Costco generated 320 comments. The number is useful. The language inside those comments is more useful.
People said they were heading to Costco. They thanked me for flagging the limited-time offer. Some named the price as the reason they were finally ready to buy.
Those comments do not prove 320 sales. They show that viewers understood the product, retailer, offer, and next action. That is a narrower claim, and it is one I can support.
Measurement discussions at Modash’s Return on Influence Festival ’26 kept returning to the same discipline. Decide what the creator asset has been hired to do, then use evidence that can address that job.
Meredith Diehn and Lucy King want evidence they can invest behind
During Ask the CMO, Meredith Diehn of goodwipes and Lucy King of Hormbles Chormbles discussed what would justify greater investment in creators.
Creator proposals are difficult to compare with media channels that sell more predictable impression ranges. The fee may cover concept development, production, access to an audience, the creator’s name and likeness, organic distribution, paid usage, exclusivity, or several assets with different jobs.
A view count cannot explain the value of that package by itself.
Meredith discussed the audience behavior inside a result. Watch depth, saves, shares, comments, clicks, and purchases describe different responses. A smaller video filled with specific product questions may help a team make a different decision from a larger video with little evidence that viewers understood the offer.
Lucy discussed the distance between creator discovery and retail purchase. A person may encounter the product in a creator’s video, search for it later, compare it elsewhere, and buy it from a store. Last-click reporting will credit the final recorded action. It may miss the creator who introduced the product or made it memorable.
Adobe assigns the metric after it assigns the job
Leah Walker described Adobe’s separation of awareness, consideration, and conversion assets.
A broad-reach creator integration may introduce a product. A tutorial or before-and-after video may answer questions for someone considering it. A short paid edit may put the feature, result, offer, and call to action in front of someone closer to purchase.
Adobe changes the opening, footage, demonstration, and edit for each objective. The report should preserve that objective too.
An awareness video should not be treated as a failed conversion ad because it did not generate enough clicks. A conversion asset should not receive a victory lap for broad attention that produced no evidence of action.
My Crest comments showed comprehension and intent
The Crest campaign had a specific retail objective. The video needed to make the limited-time Costco offer easy to understand without abandoning the comedy format people follow me for.
The campaign generated more than 1,300 engagements, including 320 comments and 93 shares. The comments let me inspect whether the commercial information survived the joke.
Some viewers repeated the product and price. Others mentioned going to Costco or being ready to buy. That language is evidence of comprehension and stated shopping intent. It is stronger than reporting “320 comments” with no account of what people discussed.
It still is not transaction data. A person can announce an intention and never purchase. Someone else can buy without commenting, clicking, or using a code. I can describe the signal without upgrading it into proof of revenue.
Comments are not one behavior
Tyler Vaught of Edelman recommended reading a creator’s recent comments while evaluating fit. I use the same level of attention after publication.
A comment can repeat the joke, ask about sizing, identify the product, tag a friend, challenge a claim, describe a prior purchase, or state plans to visit a store. A dashboard may count each one as a single engagement. The words point to different audience responses.
This does not require pretending that every comment deserves a sentiment score or a grand conclusion. It requires reading enough of the conversation to know what the count contains.
Conversion data misses purchases too
Links, codes, platform sales, and retailer data can connect a campaign with action closer to purchase. They deserve a central place in conversion reporting when the campaign supports them.
They still have blind spots. People forget a code, search for the brand separately, use another discount, buy in a physical store, or complete the purchase outside the attribution window.
The festival’s final council discussed post-purchase surveys as another source. Asking customers how they heard about the brand can reveal marketing that last-click systems overlook. The answers depend on memory and the options provided, so they should be reported with those limits intact.
Brand-lift and conversion-lift studies can provide more evidence when the campaign size and platform support a valid test. The method should be named clearly. “Lift” is not a substitute for explaining what was compared.
Creator-led and studio ads need a fair comparison
Several speakers discussed comparing creator-led ads with studio or brand-made creative in paid media.
That test is useful when the audience, placement, spend, offer, campaign period, optimization event, and attribution window are comparable. A new creator asset against an exhausted studio ad does not tell the team much about the formats.
The investment also includes what the brand bought from the creator. Production, paid usage, whitelisting, editing, extensions, and exclusivity may all affect cost. Those terms need to be defined before production, not reconstructed after the results arrive.
My report should preserve what each number can prove
For every creator asset, I want the report to keep the original objective visible beside the result. It should identify the audience and placement, separate production and rights when that distinction is useful, and name the limits of the available data.
Reach, impressions, views, completed views, and brand lift can help evaluate awareness. Watch behavior, saves, shares, comments, searches, and product questions can provide evidence of consideration. Clicks, codes, platform sales, retailer data, and controlled lift studies belong closer to conversion.
No single metric completes the story. I would rather make a precise claim from incomplete evidence than a sweeping claim the campaign cannot support.