Support tickets into a knowledge base

The output is written for customers, not a slide deck about tickets. For each batch you receive:

• A question map: tickets clustered by the underlying question, with the phrasing customers actually use and the product area each cluster belongs to.
• A gap list that compares the map against your existing help center: questions with no article, articles that exist but do not answer the question, and articles that contradict what agents tell customers.
• New deflection articles for the uncovered questions, written from the resolutions that worked, with clear steps, notes on where screenshots belong, and titles in the customer's own words.
• Rewrites of existing articles that the tickets show are failing readers.
• Structured Q&A pairs in a clean format, ready for an FAQ page, FAQ markup, or the knowledge source behind a support bot.
• Suggested macro updates for your agents, so their replies link to the new articles instead of retyping them.

Every article is traceable to the cluster it came from, so your team can see which tickets justified it and check the answer against real cases before publishing.

1. Export. You export resolved tickets from Zendesk, Zoho Desk or another helpdesk, as CSV, JSON or through the API. We send a short guide on which fields to include and which to strip before sharing.
2. Clean and cluster. We remove noise such as auto-replies, spam and one-off billing requests, then read and group the rest by the question behind them. Tooling helps with the sorting; the reading and the judgment are ours.
3. Gap analysis. We compare the clusters with your current articles and send the gap list with a proposed writing order, starting with the questions that come in most often.
4. Writing and review. We write the articles and Q&A pairs in your help center's voice and format. Your support lead reviews them for accuracy before anything is published. Articles arrive as drafts in your help center, or as pull requests if your docs live in Git, so publishing stays in your hands.
5. 30-day fix window. For 30 days after delivery, we correct anything your team finds inaccurate or unclear, at no charge.

Customer data stays under your control. We work only from the export you choose to send, sign your NDA if you have one, and delete the raw export when the project closes, as the confidentiality policy describes.

Per 500 tickets

From $1,800

about €1,650

Each batch of 500 resolved tickets is priced from $1,800. That covers cleaning, clustering, the gap analysis and the articles and Q&A pairs the batch justifies. We agree the number of batches before we start, and you can stop after any of them. The quote lists what each batch includes and what we need from your team, so there are no surprises halfway through.

What moves the price: long, multi-threaded tickets rather than short ones; tickets written in several languages; a product with many distinct areas, which produces more clusters and more articles; and whether the articles go into an existing help center or into a new structure that we also have to design. Most teams start with one batch of their most recent tickets, which shows quickly whether the rest of the backlog is worth mining.

Prices are in US dollars; the euro figures are approximate conversions for reference. If your help center is also moving platforms, the mining can run alongside a Zendesk to Docusaurus migration, so the new site launches with the missing articles already in place.