adorosario

Totally expected -- that is the best way to show businesses the value (and then start charging for it) -- there is NO way that they can scale a free tier without running ads.

I think websites with LOT more traffic than mine will start posting similar case studies soon.

A click from ChatGPT is a lot more qualified that a click from a TBC (10-blue-link)

Just to be clear: What level of data would you have liked to see?

Totally -- any outgoing link seems to be "pre-cooked" and "pre-approved" by the ChatGPT context (so intent is much higher)

I recently dug into some data in GA4 for our website and found that while Google Organic brings in more traffic, ChatGPT clicks convert way better — 6.8X better for free trial conversions, to be exact.

If you’re in the trenches of SEO and digital marketing, you know that this is a big deal.

Despite having a fraction of the total visitors, ChatGPT clicks were outperforming Google Organic in converting users into free trials (our conversion event)

From a common-sense perspective, it makes perfect sense — instead of TBC (10-Blue-Links), the user is coming to your website highly qualified and probably warmed up and pre-sold on your product and services.

In short, while ChatGPT drove just 4% of the traffic, it accounted for 22% of the conversions.

Serverless RAG: Production-ready RAG API, benchmarked at #1 for both accuracy and anti-hallucination. Never build or run your own RAG pipeline again (964 features/issues built in last 12 months)

Hi - I'm Alden from CustomGPT.ai -- what started as a simple no-code platform to build chatbots with your own data, has become one of the top serverless RAG APIs.

Since Day 1, we knew that RAG is going to be an important part of the GenAI ecosystem, due to which we focussed on accuracy and critical business features like anti-hallucination, explainability and auditability.

In recent benchmarks, we topped OpenAI Assistants and other serverless APIs like Vertex -- for both accuracy and hallucination.

Why we built this: Just like you should never have to build your own LLM (you would use OpenAI!) -- similarly, you should never have to grapple with the 964 issues (real number!) in building your own production-grade RAG pipeline.

This API lets developers build production-scale RAG applications, without dealing with any of the RAG complexities. In other words: Let someone else take care of the RAG pipeline (now and forever!)

Over the last 18 months, we've signed up over 8,000 paying customers, including marquee institutions like MIT.

We invite you to try out the API (there is a 7-day free trial). And as always, feedback in welcome!