How I scaled SEO at Wavel AI: 23 to 22,000 weekly non-branded clicks
Wavel AI is an AI SaaS platform for voice, dubbing and video. Over three years I built its SEO from a solo role into a team of 11 writers. Here's what we did and what it did to search traffic.
The short version
- Weekly non-branded clicks grew from 23 to about 22,000 (about 957×) between October 2022 and September 2025.
- In 2024, organic search brought 1.89 million users (+503%) and 1.07 million key events (+725%).
- Ahrefs Domain Rating rose from 5 to 61, and referring domains from 160 to more than 1,500.
- The SEO function grew from just me to a team of 11 writers with supporting specialists.
Wavel AI is an AI SaaS platform for voice, dubbing and video. I led its SEO from October 2022 to September 2025.
Before and after
A typical week at the start of the role, next to a typical week three years later:
- 957×Non-branded clicks a week23 → about 22,000
- 74.5×Non-branded impressions a week7,100 → 529,000
- 189×Total clicks a week233 → 44,000, branded included
| Metric | Oct 2022 | Sep 2025 |
|---|---|---|
| Non-branded clicks per week | 23 | 22,000 |
| Non-branded impressions per week | 7,100 | 529,000 |
| Total clicks per week | 233 | 44,000 |
| Ahrefs Domain Rating | 5 | 61 |
| Ahrefs referring domains | 160 | 1,500+ |
| Indexed pages | ~2,600 | 3,300+ |
Non-branded clicks are the clearest measure here: they show how far the site's visibility grew beyond people who already knew the Wavel AI name.
The situation
When I joined Wavel AI in October 2022, the site already had about 2,600 indexed pages, but very few people found it through search: 23 non-branded clicks a week, an Ahrefs Domain Rating of 5 and around 160 referring domains.
The platform had several searchable products and use cases, but the SEO work had no structure yet. Technical issues affected rendering and indexing, and every page was built by hand. My brief was to grow organic search, with room to build a team and no preset growth target.
The strategy
1. Map search demand to product clusters
I organised keyword research around the products: text-to-speech, AI dubbing, voice cloning, video translation, video-to-shorts, screen recording, AI avatars, and subtitles and transcription. I also mapped use cases across SaaS, marketing, education and accessibility, to find demand beyond the broad product keywords.
2. Expand content and language coverage
I led the creation of more than 100 pages across the product clusters: child pages, use-case pages, studio pages, solution pages and language versions. At peak, the blog published 5 to 10 posts a week, with dedicated sprints for transcript pages.
Internationally, I audited the Spanish, French, German and Arabic pages and corrected hreflang. ChatGPT and Gemini helped with prompts and optimisation suggestions, while writers and QA stayed in the publishing process.
3. Fix what stopped search engines reaching the site
Technical work ran alongside the content, including:
- JavaScript errors affecting rendering and tracking, server errors and incorrect redirects.
- Canonical inconsistencies, hreflang mismatches, pagination issues, and subdomain checks after a server migration.
- Structured data, implemented and validated across the homepage and key product pages, including breadcrumbs.
- Performance: lazy loading and asynchronous third-party scripts, plus robots.txt, HSTS and consent set-up.
This removed barriers to crawling, indexing, page speed and reliable measurement as the content library grew.
4. Keep outreach running
I ran email outreach with Brevo and mail merge, alongside guest posts, HARO submissions and link exchanges. Volumes ranged from 315 to 4,169 emails a week, with about 50 to 150 live links a month. I also supported the product's visibility on Reddit, Quora and Product Hunt.
Over the three years, referring domains grew from 160 to more than 1,500 and Domain Rating from 5 to 61.
5. Build the team and review every week
As the work grew, I built the team to 11 writers with supporting specialists. During active optimisation periods we updated 25 to 70 pages a week, and I reviewed visibility, users, sessions, click-through rate and average position every week to decide what to change next.
Organic search in 2024
The 2024 results show the scale of organic acquisition during the programme. They cover a calendar year, so they're a different period from the September 2025 weekly figures above.
| Metric | 2024 | vs 2023 |
|---|---|---|
| Organic search users | 1,888,815 | +503% |
| New users from organic search | 1,886,315 | +501.13% |
| Returning users from organic search | 215,407 | +485.49% |
| Recorded key events | 1,071,566 | +725% |
| Average user event rate | 26.42% (was 19.78%) | +6.64 pts |
Key events are the actions set up for tracking in analytics. Their count alone doesn't show how many customers were acquired or how much revenue was generated.
Weekly sessions peaked at about 65,000 in mid-2024, and the average search position improved from 35.6 in October 2022 to about 24 to 26 by late 2024.
What I learned
More coverage meant more ways to find the product
Pages built around product categories, use cases and languages widened the searches that could lead people to Wavel AI. The case rests on visibility and traffic, not on the number of pages published.
Technical work held the programme up
Rendering, canonical, hreflang and server issues needed attention as the site grew. Fixing them gave publishing and optimisation a dependable base.
Link metrics need careful reading
Outreach ran alongside content and technical work. The backlink profile grew a lot, but the data can't isolate how much of the traffic growth came from links. I'd judge that work by page performance, not treat Domain Rating as proof of SEO success.
A repeatable process made scale possible
The lasting change was going from one SEO to a team that could publish, review and improve content consistently. Writers, QA, technical fixes and weekly reporting gave the programme a process that kept working as it grew.
How it's measured
- Google Search Console · weekly clicks and impressions, split by branded and non-branded queries
- GA4 · Organic Search · users, key events and event rate, 2024 vs 2023
- Ahrefs · Domain Rating and referring domains
Weekly figures compare my first week (October 2022) with a typical week in September 2025; 22,000 and 44,000 are approximate.