Paid knowledge platform
A paid knowledge platform makes a fifteen year archive worth paying for again
Subscribers used the newest articles and ignored the rest, which made the archive hard to justify at renewal. Semantic answers across three subscription tiers changed which content gets read.
Problem
The platform sold access to nine thousand specialist articles on Drupal, across three subscription tiers. Analytics showed a familiar pattern: subscribers read the last few weeks of publishing and almost nothing older, even though the archive was the reason the subscription cost what it did. At renewal time, customers judged the product on the handful of recent articles they happened to read, and the depth they were actually paying for was invisible to them.
Solution
One index covers the whole archive, with the subscription tier applied as a filter before retrieval rather than as an instruction to the model. A basic subscriber asking a question that is best answered by a professional tier article gets a correct answer from the content they do have access to, plus a clear note that deeper material exists at another tier. Publication dates are carried into every answer, and articles superseded by later ones are down ranked so the chatbot does not quote guidance that has since changed.
Result
Sessions per subscriber rose noticeably, and the change came almost entirely from older material being surfaced by questions rather than found by browsing. The upgrade prompt turned out to be the most direct commercial effect: subscribers now regularly discover that the answer they want sits one tier up, at the exact moment they need it.
Impact
The archive went from a storage cost to the strongest argument for renewing. Nothing was rewritten and no article was republished. The content had always been good enough, it just was not reachable by anyone who did not already know it existed.
Could this be your knowledge base?
We will show you what it looks like on your own content.