/ Case study · 5 min
Case Study: We Followed Our Own Report, and Our Home Page Went from 79 to 98
We ran a VISIBLA Full Report on our own home page, made the five changes it asked for, and checked the live page again. Here is what we changed, what each change was worth, and what we skipped.
By the VISIBLA team · Published
Why we audited our own website
Before we ask you to trust a score, we should show what happens when we follow our own advice. On November 3, 2026, we ran a Full Report on our home page, visibla.app, and did what it said.
When the audit asked how customers get our service, we answered “Online, for customers anywhere”, because that is what VISIBLA is. That answer decides which contact and location checks apply to the page.
Where we started: 79 out of 100
The report scored the page 79, labeled Good. No critical gate limited the score, but points were missing in six of the seven categories:
| Category | Before | After | Out of |
|---|---|---|---|
| Service Clarity | 15 | 20 | 20 |
| Search Intent Match | 12 | 15 | 15 |
| Local Signals | 12 | 15 | 15 |
| Trust Signals | 13 | 13 | 15 |
| Conversion Path | 14 | 15 | 15 |
| Structured Data | 5 | 10 | 10 |
| AI Access & Answerability | 8 | 10 | 10 |
| Total | 79 | 98 | 100 |
Our weakest area was Structured Data: the code behind the page that tells machines what a business is. Ours described the page only as a web application. It said nothing about the company behind it, the service it offers or the questions it answers.
The five changes we made
The report lists its fixes in the order that moves the score most. We made five of them, kept every change true, and skipped two. The running score below is the report’s estimate after each step; the final 98 is the new check of the live page.
| Change | What we did | Score |
|---|---|---|
| 1. Name the service in the main heading | Moved our existing line “AI Visibility for U.S. service businesses” into the main heading. | 84 (+5) |
| 2. Answer buyer questions | Added six common questions to the page, each with a short, direct answer right below it. | 88 (+4) |
| 3. Say how fast we reply | Added “Response time: we reply within one business day.” | 92 (+4) |
| 4. Show a real update date | Added a visible “Last updated” date, and the same date in the page code. | 93 (+1) |
| 5. Describe the business in the page code | Added structured data for our company, our service, the page and its questions, matching what the page shows. | 98 (+5) |
1. Name the service, not just the slogan
Our headline was “Get found by AI. Get chosen by humans.” It is memorable, but it doesn’t say what we sell. The line that did, “AI Visibility for U.S. service businesses”, sat just above it as a small label, outside the heading. We moved that same line into the heading. The page looks the same, and the heading now names the service.
2. Put the answers where people can see them
We already had a full FAQ page, but the home page only linked to it. We added six questions to the home page itself: who VISIBLA is for, what the score means, what the free audit includes, whether a high score guarantees AI recommendations, who is behind VISIBLA and how fast we reply.
Each answer sits directly under its question and is always visible, not hidden behind a click. An AI tool that wants to quote a short answer needs the question and the answer next to each other.
3. Make a promise you can keep
The report asked how fast we reply. We checked what we could actually commit to and wrote it plainly: we reply within one business day. The same sentence is on the home page, the FAQ page and the contact page, so they never disagree.
4. Date the page honestly
We added a “Last updated” line under the questions, with the date we reviewed the content, and put the same date in the page code. We’ll change it when we review the page again, not on a timer. A date that changes while the content stays the same tells readers nothing.
5. Make the page code match the page
Behind the page, we now describe the company (Briviom LLC, which makes VISIBLA), the service (AI visibility audits for U.S. businesses, in English and Spanish), the page itself with its update date, and the same six questions and answers that are visible on it. Every fact in the code is also on the page.
What we skipped, and why
- Review markup. The report suggested matching review data to visible reviews. There are no customer reviews on our page yet, so there is nothing true to mark up. Review markup without real reviews would be misleading.
- Breadcrumb markup. A home page sits at the top of the site, so it has no breadcrumb trail to describe. We left it out, and the page still reached the full 10 points for Structured Data.
Where we ended: 98 out of 100
The changes went live on November 3, 2026. We then checked the live page again with the same methodology, version 2026.11. It scored 98: full marks in six categories and 13 out of 15 in Trust Signals.
Our Trust Signals score is the one we earned least. Under the corrected rules, described below, ordinary words such as “reviews” no longer count as proof. The Trust points now come from this case study, which the page shows under its own “Case study” heading. The 2 missing points need genuine customer reviews, shown on the page and matched in the code. We are collecting feedback from our customers now, and we’ll publish it only with their permission.
A flaw we found in our own scoring
Writing this case study showed us a flaw in our own scoring: ordinary words such as “reviews” could count as proof. We corrected methodology v2026.11 on November 3, 2026. Re-scored with the corrected rules, our old page drops from 79 to 74, and the new page still scores 98.
What this score does and doesn’t mean
A score of 98 means the page is clear: it says what we do, who it’s for, how to reach us and when we’ll reply, in a way people and AI tools can read. It doesn’t mean any AI tool will recommend VISIBLA, and no honest score can promise that. How the score works →
What you can take from this
- Most of the gain came from saying clearly what the page already knew: the service, the reply time and the answers to common questions.
- None of the changes needed a redesign. The page looks almost the same.
- Every change can be checked. Each one is either visible on the page or matches something that is.
FAQ
Did you change the scoring to get a better result?
We changed it, but not to get a better result. The correction made scoring stricter. It lowered our old page from 79 to 74, and the new page still scores 98.
Will my page get the same score if I make the same changes?
Not necessarily. Every page starts from a different place, and some checks depend on how customers get your service. Your report lists the changes that matter most for your page, in order.
Why not 100?
The last 2 points need genuine customer reviews on the page. We would rather show 98 than fake them.