Guna

Meet ChatRAG your AI support agent that actually understands your content. 📄 Upload PDFs, Docs, or URLs 🎨 Customize personality & tone 🌐 Embed anywhere, instantly Start building smarter support: chatrag.co

Day 38 of the ChatRAG SEO experiment. Good news and still waiting news. ✅ New tool page indexed AI Support Readiness Checker is live on Google ❌ BOF page still "Discovered – currently not indexed", Last crawl N/A ❌ Learn page still "Discovered – currently not indexed", Last crawl N/A The tool page indexing confirms domain trust is rising from the backlink boost. Today's fix: added a contextual internal link from an already-indexed learn page to the stuck learn page. The stuck learn page only had sitemap as its referring source no internal links from indexed content. That's likely why Google hasn't prioritised crawling it. A link from an indexed page gives Googlebot a path to follow. Sitemap alone isn't enough on a low-trust domain. BOF page nothing to touch. The DR 66 and DR 71 backlinks are in place. Waiting for Google to act on them.

Day 37 of the ChatRAG SEO experiment. BOF page: still "Discovered – currently not indexed" Last crawl still N/A. But today I checked the backlink profile on Ahrefs. Here's what 37 days of building in public actually built: → 88 backlinks → 22 linking websites → 77% dofollow → DR 66 from Aura++ → DR 71 from TinyLaunch → DR 63 from TrustMRR A press release this week added a significant boost on top of the existing profile. Domain Rating is still 1.1 DR lags behind actual backlink acquisition by weeks. But the trust signals are now real. High authority dofollow links pointing at the domain is exactly what Google needs to start allocating crawl budget to stuck commercial pages. Expecting Last crawl to change from N/A within the next 7 days.

Day 36 of the ChatRAG SEO experiment. Shifted from content to engineering as marketing today. Built a new free tool: AI Support Readiness Checker Paste your docs or a URL → get: → A readiness score out of 100 → Estimated ticket deflection % → Topics your docs cover well → Gaps that would limit an AI agent → Top recommendation to improve The insight behind it: most founders don't know if their documentation is actually good enough to power an AI support agent until they try it and get wrong answers. This tool answers that question before they deploy anything.

Right now I’m targeting founders / small teams who already have docs and are considering AI support but aren’t sure if they’re ready yet. Distribution-wise, I’m treating this as an entry point sharing it where people complain about bad docs or AI answers (X, Reddit), and using it as a way to start conversations rather than just drive traffic. Still early though figuring out where it actually resonates.

@gunaa_dev That sounds awesome! Who would be your target audience for this and howre you planning to get them to try it out?

@indie_saas Appreciate that Dave, seriously 🙏 yeah that's exactly the mindset shift stop guessing, start listening. emails went out this morning so waiting on replies. even 3-4 honest ones will tell me more than 7 months of assumptions. will keep posting updates as they come in 🫡

@gunaa_dev It may not seem like it Guna but its progress. If you can get any of those people to provide some insights its gold. I think you are right, you just need to get to talk to them and try to pivot around what they would consider paying for

7 months building ChatRAG. 38 signups. 0 active users. 0 paying customers. Today I'm doing something uncomfortable. 🧵

@_zidkim Means a lot man, genuinely 🙏 yeah this is just the startq going full build in public from here, wins and losses both. email blast went out this morning to all 38 signups, waiting to see who responds. next update dropping soon

@gunaa_dev ❤️🔥 This is an awesome post my friend. Appreciate the vulnerability as well as the clear drive that you have in pressing forward and figuring out how to move the needle. Will you be doing more of these posts?

@49agents honestly both but the drop-off on complex questions is something i'm still digging into. most of my 38 never even got to that stage though, which tells me the friction was even earlier onboarding or just not seeing the value fast enough to try

@gunaa_dev nobody stayed sounds harsh but its useful signal. the problem with most AI support agents is they solve the easy 10% really well and hit a wall on everything else. were you seeing drop-off on complex questions or just low engagement overall

7 months in. Not giving up The first paying customer is out there. I just need to find the right problem to solve for the right person. Let's go. 🫡

And I'm also doing something I should've done from day 1: Offering to personally set up their AI support agent. For free. If they see it working on their own site, they'll know if it's useful. If not, I'll know what to fix.

Day 35 of the ChatRAG SEO experiment. BOF page: still "Discovered – currently not indexed" Last crawl still N/A. But today I fixed something that was likely blocking the new learn pages from indexing. The problem: Every learn page component had "use client" at the top + a session fetch on mount. This was causing Next.js to bail out of server-side rendering triggering a "BAILOUT_TO_CLIENT_SIDE_RENDERING" warning in the page source. The fix: → Moved the session fetch into the header component itself → Removed "use client" from all 6 learn page components → Learn pages are now proper server components → Default session state is false no more loading spinner Google can now read these pages without executing JavaScript. Expecting the learn pages to index within 24 hours.

Day 34 of the ChatRAG SEO experiment. BOF page: still "Discovered – currently not indexed" Last crawl still N/A. Today I did a deep technical check on why the new learn pages aren't indexing yet. Checked the page source directly. Good news: full page content IS server-side rendered. H1, paragraphs, sections all visible in the HTML. Google can read it. Found one thing worth noting: The "use client" directive in Next.js components is triggering a "BAILOUT_TO_CLIENT_SIDE_RENDERING" warning in the source. Not a blocking issue content is still readable. But something to clean up. Also fixed two 404s yesterday: → /auth/signup → redirected to /auth/register → /chat-iframe/ → redirected to homepage Domain is healthy. Content is readable. Waiting for the crawl.

Day 33 of the ChatRAG SEO experiment. BOF page: still "Discovered – currently not indexed" Last crawl still N/A. Today was a deep dive into GSC data. Here's what I found: Domain health: → 28 pages indexed and growing → 3.89K total crawl requests Google IS crawling actively → 93% OK (200) responses site is healthy → www redirect working correctly Issues found and fixed: → /auth/signup returning 404 redirect added to /auth/register → /chat-iframe/ returning 404 redirect added to homepage → CSS chunk 404s old Next.js build files, unfixable but harmless The BOF page situation is now clearly not a structural problem. Domain is healthy. Crawl budget is active. Internal links are in place. External signals planted. It's purely a trust and time issue at this point. Experiment continues.