Strong domain, off-target traffic, and the category query AI hands to competitors

SEO Audit: TurboStarter

TurboStarter is a SaaS starter kit, not just a boilerplate. It gives founders a full stack to ship an AI-first web app, mobile app, and browser extension from one codebase. The product is validated, it already makes sales, and a lot of SEO work is clearly in place.

But it plays in a brutal market. ShipFast, MakerKit, and supastarter are all fighting for the same searches, most of them built and distributed by sharp solo founders. In this audit we go through Ahrefs, Google Search Console, the technical setup, backlinks, and AI visibility. Two things stand out: almost all the traffic comes from a single page that has nothing to do with the product, and a wave of off-topic AI pages is diluting an otherwise strong domain.

Current performance

That traffic is deceptive. 97.7% of it lands on the 2FA documentation page, which ranks position 11 for "2fa auth" (46K volume). Geography tells the story: 79.5% Cambodia, 13.9% Bangladesh, 1.1% United States. The backlink profile is healthy and natural with a clear directory strategy already running — no damage from the April core update — and the brand is starting to show up in AI Overviews and ChatGPT at a small scale.

The core problem: the strong DR and backlink profile are not converting into qualified, on-topic traffic. The one page that ranks (2FA docs) pulls cold developer traffic from Southeast Asia with no buying intent, while the pages that should sell — the starter kit, the stack comparisons — rank for almost nothing.

Finding #1: The traffic is real but off-target

That 2FA documentation page is a quiet asset. The docs are written well enough to rank for a 46K-volume term and pull over a thousand visits a month. The problem is intent: these are developers looking up how to set up two-factor auth, mostly outside your target market, not founders shopping for a starter kit.

Do not delete it. Use it. Add contextual CTAs and internal links from the high-traffic docs back to the product and the comparison pages, and treat that audience as top-of-funnel to capture into email. It is cold traffic, but it is developer traffic — and some of them ship SaaS.

Finding #2: Prune the off-topic AI pages

The sitemap lists roughly 3,000 pages, but Google has only indexed about 800. The gap is a large set of AI-generated “ideas” pages (think “AutoDrop Genie,” a drop-shipping concept) that have plenty of copy but sit completely outside TurboStarter's semantic field.

Search Console makes the cost visible. The site has 922K impressions over three months but a 0.2% CTR. One idea page (“serpinsight”) alone pulls 338,949 impressions and converts just 49 clicks. Google is surfacing these pages, users are not clicking, and the whole domain absorbs the weak engagement signal.

The move is to prune. Keep what supports the business (docs, stacks, comparison pages, real tools) and remove or noindex the off-topic idea pages. A tighter, focused site sends a far stronger quality signal than a bloated one, and it stops crawl budget leaking into URLs that will never sell.

Finding #3: Win one primary keyword

The homepage spreads across “AI-first SaaS starter” and “boilerplate” at once. Pick one primary keyword and commit to it. Based on the data, “SaaS starter” is the strongest candidate. Then make the page reflect it:

• Meta description: add “SaaS starter” and “boilerplate.” They are missing today. Keep it under 150 characters.
• H1: “Ship a production SaaS this weekend” reads well but does not carry the keyword. Move it to an <h2> for conversion, and promote a keyword-led line (e.g. “Best AI SaaS starter kit”) to the real <h1>.
• H2: make sure at least one subheading contains the exact target keyword.

The page is already well built and A/B tested for conversion, so these are surgical on-page edits, not a redesign.

Finding #4: Build the comparison and stack cluster

This is where the competitor data is impossible to ignore. MakerKit ranks for 160 keywords against TurboStarter's 20, and it wins them on exactly the terms your buyers research: "supabase pricing", "drizzle vs prisma", "nextjs hosting", "cloudflare workers pricing". Same stack, same audience, far more coverage.

The versus pages (TurboStarter vs ShipFast, vs MakerKit, vs supastarter) are excellent and already interlinked. Now go deeper:

1. 1One page per stack comparison. Build dedicated pages for “Drizzle vs Prisma,” “Stripe vs Lemon Squeezy,” “Supabase vs Neon,” and so on. Cluster them around hubs (payments, database, hosting) with exact-match internal anchors.
2. 2One versus page per competitor that ranks for “best SaaS boilerplate,” not just the four you have today.
3. 3Publish your own listicle. A “best SaaS boilerplate” roundup that lists every competitor, with TurboStarter first. Listing your own brand at the top is one of the strongest signals for AI citation right now.
4. 4Get on the lists that already rank. Pages like the lovable.dev roundup pull AI citations. Reach out, make the case for inclusion, and earn the backlink at the same time.

Finding #5: Mine Search Console for intent

Average position is a healthy 8.3, so the foundation is there. Two strategies pull more from it:

• Question queries. Filter your queries with a regex for interrogatives (who, what, how, vs, difference). These long-tail, high-intent terms are the same prompts people type into ChatGPT and Claude, which makes them the best queries to own for both classic and AI search.
• High impressions, low clicks. These pages already surface but are not optimized for the query. “Indie hackers discord” pulls 1,200 impressions and only 20 clicks, pointing straight at your community page. A title, meta, and heading refresh is often enough to convert impressions into clicks.

Finding #6: Free tools as a channel

There is one free tool live (the SaaS ID generator), and it already ranks. That is proof the format works, so build more. Aim for roughly two per month, grouped under a tools index page with child pages beneath it and breadcrumb linking both ways. Competitor keywords are a good idea source — ShipFast ranks for a logo generator, for example.

Quick technical fixes

• Mobile performance is 33/100. Largest Contentful Paint on the homepage is 17.3 seconds and Total Blocking Time is over 3 seconds, held up by render-blocking requests, legacy JavaScript, and a heavy CSS chunk above the fold. CLS is a clean 0. The LCP and blocking fixes are fast, cheap wins for both UX and crawlability.
• Canonical tags on docs. Some pages (e.g. the mobile docs) point their canonical to the web docs instead of themselves. Unless that is deliberate, make them self-referencing. A free crawl with Screaming Frog will surface the rest.

AI visibility

The platform is already set up for AI: every retrieval crawler is allowed, the site is server-rendered, and Claude already cites TurboStarter on direct-brand queries. In our live snapshot:

The verdict

TurboStarter has a strong foundation: high DR, real sales, a clean backlink profile, an AI-ready platform, and a genuine product moat that Claude already recognizes. The work now is focus and intent.

Do that, and TurboStarter stops being the boilerplate AI forgets in the category search and becomes the one it recommends first.