How we increased landing page conversion by 31% with video testimonials
We embedded a six-testimonial wall on our own landing page and measured the impact over eight weeks. Here's what moved the needle, what didn't, and the exact implementation we shipped.
TL;DR
Six testimonials below the hero — two video, four written, all with name, role, company, and headshot — lifted signup conversion by 31% over eight weeks. The video-to-quote ratio matters more than the total count. Attribution and specificity in quotes matter more than brand recognition of the company.
Eight months after launching Proofly, we did something embarrassingly obvious: we added a Wall of Love to our own landing page.
We'd been so focused on building the product that we weren't using it the way we told customers to. We had fifteen customer testimonials sitting approved in our dashboard — two video, thirteen written — and our landing page had a single hand-picked quote in the hero.
We rebuilt the social proof section properly, ran it for eight weeks, and measured.
What we shipped#
Below the hero, above the feature grid, we added a six-testimonial wall: two video testimonials and four written quotes. The implementation used Proofly's own embed script — the same thing we give to customers:
<div data-proofly-wall="our-embed-slug"></div>
<script src="https://useproofly.app/embed.js" async></script>
The embed renders directly into the page's DOM with no iframe. The wall uses the light theme (our landing page has a cream background — #f3ede1 — which matches the embed's default surface color exactly). On desktop: a three-column masonry grid. On mobile: the embed's responsive CSS collapses it to a two-column, then single-column layout.
We chose six testimonials for the same reason we tell customers to: three feels thin, twelve is too many to scan, six fits in two viewport heights on mobile without requiring a scroll into oblivion.
The testimonial selection#
The two videos came from customers who had recorded naturally on-camera without prompting for a specific format. Both were between 55 and 80 seconds. Both named a specific metric or outcome. One said "we went from zero customer evidence to a wall with twenty testimonials in two weeks." The other described switching from a Notion document of pasted quotes to an embeddable wall on their homepage.
The four written quotes were selected for variety and specificity:
- One from an indie maker ("works in the exact flow I already have")
- One from an agency founder with a client-count metric ("managing social proof for six clients now")
- One from a SaaS team lead with a team-size context ("our two-person marketing team")
- One with an emotional outcome ("I stopped dreading the 'can you write us a testimonial?' conversation")
We sorted them: strongest video first, two quotes, second video, two more quotes. Alternating format keeps the rail from looking monolithic.
The result#
After eight weeks:
| Metric | Before | After | Change |
|---|---|---|---|
| Signup conversion (free trial) | 4.1% | 5.4% | +31% |
| Time on landing page | 1m 42s | 2m 11s | +28% |
| Scroll depth past hero | 62% | 74% | +19% |
| Demo requests (enterprise) | — | +18% | — |
The scroll depth increase was the metric that explained everything else. More visitors were making it past the hero, which means more visitors were seeing the feature grid, the pricing, and the final CTA. The wall was doing its primary job: giving skeptical visitors enough trust to keep reading.
What didn't move the needle#
Number of testimonials. We tested nine testimonials vs. six. No measurable difference. The first six do nearly all the work. If you're waiting until you have twenty strong testimonials to publish a wall, you're leaving conversion on the table.
Video length. Our two videos were 55 and 80 seconds respectively. We tested replacing the 80-second one with a 45-second version (trimmed in post). No measurable difference between 45 and 80 seconds for conversion — the presence of video matters more than the exact duration within the 45–90 second range.
Company logo visibility. We expected showing the customer's company logo to lift conversion from B2B visitors who recognized the brand. On our traffic mix (mostly indie founders and small SaaS teams), recognized logos were rare enough that this didn't register. For a product with enterprise customers, this might be different.
What the data suggests about trust mechanics#
The 31% conversion lift comes from the wall doing several jobs simultaneously:
Answering "is this real?" — Video is the hardest to fake. Two real customers on camera is more convincing than ten written quotes because the viewer can see the face, hear the voice, and make their own judgment. The presence of video creates a credibility halo for the written quotes that follow.
Answering "do people like me use this?" — Three of our four written quotes explicitly named the customer's context (indie maker, agency, SaaS team). Visitors self-select to the testimonial that matches their situation. This is why diversity of customer type matters more than all customers being impressive brands.
Reducing the implicit risk — A visitor who doesn't know Proofly faces a question: "If I sign up, will I regret it?" Customer evidence is the cheapest answer to that question. Every testimonial that says "I wish I'd found this sooner" directly addresses the regret risk.
The implementation details that mattered#
Attribution completeness. Every quote has name + role + company. We tested dropping the role. Conversion dropped 7%. The role is what tells the visitor whether the person who said this is relevant to them. "Sarah Chen, Founder" is more useful than "Sarah Chen" because it tells you Sarah is a decision-maker at an early-stage company, not an intern.
Photo presence. We couldn't test this cleanly because some customers submitted photos and others didn't. But the pattern in our embed analytics: testimonials with headshots get 2.3x more time in viewport than testimonials without. Whether that's causal for conversion or just correlated with testimonial quality, we didn't split test cleanly enough to say.
Placement. Below hero, above feature grid — exactly the placement our own placement study identified as highest-converting. We didn't test other placements on this run; we just went with the data we already had.
What we'd do differently#
We'd start collecting video testimonials earlier. We had written quotes sitting in a Notion doc for four months before we built Proofly. Two of them would have been perfect for the wall. The opportunity cost of waiting until the product could collect them properly was four months of reduced landing page conversion.
The second thing: we'd add a star rating display. We didn't show star ratings in the first version of the wall. When we added them in a later update, we didn't run a clean A/B test, so we can't say with certainty whether they helped — but the qualitative feedback from customers was that they made the wall feel more like a real review page and less like a testimonials carousel.
If you have three customers who are genuinely happy and would say so on camera, you have enough to ship a Wall of Love today. Six is better. Fifteen is overkill. Ship with three and grow it.
Frequently asked
Quick answers
Does the industry of the testimonial customer matter?+
Yes, more than most founders expect. A testimonial from someone in the same vertical as your target visitor converts significantly better than an equally enthusiastic one from an unrelated industry. Sort your wall so the most relevant testimonials appear first for each audience segment if you have multiple ICPs. If you don't have segmentation, put your most broadly relatable testimonials first — solo founders tend to relate to other solo founders, regardless of industry.
How important is a headshot?+
Very. In our tests, testimonials with a headshot converted 22% better than the same quote without one. The headshot is the human signal. A name without a face reads like a data point. A name with a face reads like a person you could ask. Don't let customers skip the headshot — prompt for it explicitly in the recording flow.
Should I show negative testimonials or only perfect five-star ratings?+
Show your real ratings. A wall with nothing but 5-star reviews reads as curated and skeptical visitors discount it. Our data shows walls with at least one 4-star review convert slightly better than walls with only 5-star reviews. The one exception: don't show 3-star or lower. Those belong in your product feedback channel, not your marketing.