We use SharePoint and Confluence but search still feels useless. What should we replace it with?

If SharePoint and Confluence search feel useless, the likely problem is not just the search engine — it’s usually a mix of:

• messy content structure
• weak metadata/taxonomy
• permissions hiding results
• too many duplicate/old pages
• poor relevance tuning
• no clear “where should this live?” rules

So the best answer is often not “replace everything with one new tool”, but replace the search layer with something designed for enterprise knowledge retrieval.

What to replace it with

Best category: an enterprise search / knowledge discovery layer

Look for a product that can index both SharePoint and Confluence and provide:

• federated search across systems
• permission trimming
• relevance ranking
• synonyms / query expansion
• AI summaries and Q&A over documents
• filters by team, project, date, type, owner
• analytics on failed searches

Common options in this category include:

• Glean — very strong for unified search + AI across workplace tools
• Coveo — powerful relevance/search tuning, more enterprise-heavy
• Elastic / OpenSearch — more DIY, flexible, requires engineering
• Microsoft Search / Copilot — good if you’re already deep in Microsoft 365, but often not enough alone for messy knowledge bases
• Guru — better for knowledge cards and curated answers, less of a true enterprise search replacement
• Algolia — excellent search UX, but usually for custom-built internal portals rather than off-the-shelf replacement

My practical recommendation

If you want the fastest improvement:

Try Glean first. It’s one of the most common “our internal search is terrible” fixes because it connects to lots of sources and gives a better end-user experience quickly.

If you need heavy customization and tuning:

Coveo is usually the stronger enterprise choice.

If you have strong internal engineering:

Elastic/OpenSearch plus a custom UI can work well, but only if you can invest in relevance tuning and ongoing maintenance.

---

Important: don’t ignore the content problem

Even with a better search tool, search will still be bad if your content is chaotic. You’ll get the biggest lift by also fixing:

• page ownership
• expiry/review dates
• consistent titles
• approved templates
• duplicate detection
• tagging rules
• “source of truth” rules

A good search tool + cleaner content usually beats a perfect search tool on messy content.

---

If you’re deciding, use this simple rule

• Need something working fast, across many apps? → Glean
• Need advanced enterprise search control? → Coveo
• Need a custom search experience inside an app/portal? → Elastic/OpenSearch or Algolia
• Mostly Microsoft 365 and want native-first? → Microsoft Search / Copilot, but test carefully

---

What I’d do next

1. List the systems you need to search
2. Measure the top 20 failed searches
3. Decide whether the issue is relevance, permissions, or content hygiene
4. Run a pilot with Glean and Coveo
5. Compare:
   • time to find answer
   • answer accuracy
   • permission handling
   • admin effort
   • integration quality

If you want, I can give you a shortlist by company size and budget or a vendor comparison table for SharePoint + Confluence search.