HoldFast vs System1 Test Your Ad

System1 predicts market effect against a very large norm base. HoldFast tells you which of your two versions did better on the question you chose. Knowing which question you are asking decides the tool.

System1 Test Your Ad measures real consumer emotion to predict short-term sales potential and long-term brand growth. HoldFast audience testing shows two versions of your work to two fresh groups and reports which one did better on one question you chose. Those are different questions, and the useful comparison is being clear about which one you are asking.

What System1 measures

System1 gives an ad a Star rating, a Spike rating and a Fluency rating, using an emotion measurement method it calls FaceTrace, and benchmarks the result against a database it describes as 120,000+ ads. It tests for TV, digital, social, outdoor, audio and print. Results come from 24 hours. The tiers are named Essential, Pro, Pro+, Screen, Guidance and Competitive Edge, there is no public price, and the site asks you to contact them to review your testing needs. It is built for brands and agencies who want to know, before a media spend, how an ad is likely to perform in the market.

The different question

System1 answers: how will this ad do, compared with a very large body of ads that came before it? The output is a prediction of market effect, made possible by the size of the norm base.

HoldFast answers: which of these two versions did better with this audience on this question? Every participant sees exactly one version on a plain confidential grey page, watches it once with no skipping ahead or rewatching, then answers a short survey. Only people who finished count. The question that decides the verdict is chosen before launch and frozen. Question sets are organised by funnel stage, attention, comprehension, emotion, brand, intent and fit, 21 of them, plus a brand lift study with a no-exposure control group. Video, image and audio can be tested.

The verdict is worded to match what the numbers can carry. One version outperformed the other only when every version reached the sample size committed at launch and drop-off was balanced. Below that, a directional advantage. Inside what chance could produce, too close to call, framed as information rather than failure.

Screenshot: testing-report-bottomline-1
A HoldFast report opens with the verdict in plain sentences, not a score against a norm.

What HoldFast does not have

Plainly: HoldFast has no norms database on anything like that scale, and makes no claim to predict sales. It cannot tell you how your ad compares with the category's history. It can tell you, in a written report that opens with The Bottom Line, continues with real attributed participant comments in In Their Words, and finishes with By the Numbers, which of your two versions landed better and why people said so.

Cost and door

System1 is bought through a conversation. HoldFast is bought in the wizard: a short recruited test runs about 4 credits per person, one credit being one dollar, so 24 people is around 100 credits and 200 around 800, with a floor of 25 credits on a paid test. A longer piece or a narrower audience costs more per person. The exact price is on screen before launch, nothing is charged until you approve, and unfilled seats are refunded at settle.

Who should pick which

Pick System1 when the decision is a media spend and the question is how this ad will perform against everything that came before it, across TV, digital, social, outdoor, audio or print, and a benchmarked prediction is what the room needs.

Pick HoldFast when you have two versions of one piece, one argument about which is stronger, and a client meeting where the answer has to be plain and defensible. The same upload serves the test, the review link and the showreel page, so the report travels with the cut you send.