AdRoastAdRoast/TeardownsRoast your ads free
← AdRoast Teardowns

Why does Silverfort's funnel work against itself?

S
Silverfortsilverfort.com · July 27, 2026
CybersecurityIdentity securityRuntime enforcementMessage sequencing
Out of orderDiagnosis
The short version

Silverfort has solid raw material arranged in the wrong order. The ad body does one smart thing, it names Mythos and runtime enforcement in the same beat, a frame a CISO recognizes. Then the image throws it away with vague superlatives, and the page keeps its most convincing copy farthest from first read. This is not a copy quality problem, it is a sequencing problem. When the strongest message shows up first, the buyer leans in.

The original teardown, on LinkedIn

The full infographic and commentary, straight from the post. Open on LinkedIn

The ad to landing page journey

The ad
Ad body names Mythos and runtime enforcement: a frame a CISO recognizes
Image: "Attacks stopped. Identities protected.", could be any identity vendor
"...the most advanced AI meets the strongest identity defense": drama, not proof
Image leads with a superlative
→click
The landing page
Good H1, plus a strong sub-pillar: service accounts, NTLM, legacy apps, OT
FAQ #1 answers the practical rip-and-replace fear, the best asset on the page
The strongest copy sits farthest from first read; an anonymous testimonial adds fear
Widens into 8 sub-products

The strongest message is buried below the fold.

👍

What the ad gets right

  • The ad body names Mythos and runtime enforcement together, giving the click a frame a CISO can recognize.
  • The landing page gets stronger on scroll: a good H1 and a concrete sub-pillar.
  • FAQ #1 is the best asset, because it answers the buyer's real fear around rip and replace.
🔥

Where it breaks after the click

  • The ad image uses category language instead of ICP language.
  • The hero leads with abstraction while the proof sits lower.
  • The anonymous testimonial raises over-blocking fear instead of reducing it.
  • The bottom of the page widens into 8 sub-products and dilutes runtime control.

The fixes I would prioritize

1

Remove the anonymous testimonial.

2

Build the post-Mythos version now, before the news hook fades.

3

Add the homepage logo strip: GM, Singtel, Rio Tinto, RWC.

4

Rewrite the ad image with SAP, NTLM, and service-account language.

“When the strongest message shows up first, the buyer leans in. When it gets buried, paid traffic has to work harder than it should.”
Marina Kogan, on the Silverfort funnel
Source

This teardown is based on Marina Kogan's public analysis of Silverfort's live LinkedIn ad and landing page, published July 27, 2026.

View the original LinkedIn post

Questions, answered

What does Silverfort's ad get right?

The ad body names Mythos and runtime enforcement together, giving the click a frame a CISO can recognize. The landing page gets stronger on scroll: a good H1 and a concrete sub-pillar. FAQ #1 is the best asset, because it answers the buyer's real fear around rip and replace.

Where does the Silverfort funnel break?

The ad image uses category language instead of ICP language. The hero leads with abstraction while the proof sits lower. The anonymous testimonial raises over-blocking fear instead of reducing it. The bottom of the page widens into 8 sub-products and dilutes runtime control.

How would you fix the Silverfort funnel?

Remove the anonymous testimonial. Build the post-Mythos version now, before the news hook fades. Add the homepage logo strip: GM, Singtel, Rio Tinto, RWC. Rewrite the ad image with SAP, NTLM, and service-account language.

Follow the series

Get these teardowns in your Google feed. Tap the button to make AdRoast a preferred source in Search, Discover, and News.

Want this on your own ads?

AdRoast pulls your live LinkedIn and Google ads, scores each one against your buyer, and shows exactly where the click leaks. Paste a URL, that is it.

Roast your ads free →