Marketplace search results and rankings

Marketplace homepage

The Marketplace homepage has several sections enabling customers to discover apps. The primary navigation equips the customers to find apps based on products, categories, and collections. Each product page includes apps for the respective products and customers can further refine their search using filters. The collections also act as another avenue for customers to explore apps that have performed well based on different parameters.

How search ranking works

Over the first half of 2026, Atlassian Marketplace has migrated its search backend from Algolia to OpenSearch. Unlike Algolia, which works purely on keyword-matching and has generic ranking capabilities, OpenSearch has richer ranking capabilities, supports natural language queries, and also relies on semantic matching. OpenSearch uses three factors to rank apps:

1. Keyword relevance

How well the terms in a customer's search query match the text in your app listing. Best practices:

• Identify the customer problems your app solves and include relevant keywords — especially in the app name, tagline, and brief description.
• Avoid keyword stuffing. Including irrelevant keywords makes it harder for customers to understand your app's value and can negatively impact your semantic ranking.

2. Semantic match

OpenSearch uses machine learning and Natural Language Processing (NLP) to understand the meaning behind a search query — not just the words. It matches listings to customer needs even when different terms are used. Semantic matching is based on vectors generated from your:

• App name
• App key
• Developer name
• Tagline For example, if a customer searches for "tools for timesheet tracking," OpenSearch doesn't require the word "timesheet" to appear verbatim. It understands the business need and surfaces relevant apps. Best practice:
• Write your listing — especially the description — in natural language that reflects how customers talk about their needs and the problems they're trying to solve.

3. Engagement score

Measures the level of engagement your app has received on the Marketplace. Factors include:

• Number of ratings and reviews
• Average app rating
• Number of installs
• Whether support details are provided Best practices:
• Encourage customers to rate and review your app. More reviews and a higher rating improve your engagement score and search rank.
• Add screenshots, videos, and — if possible — a demo. Rich media drives engagement and trial starts.

What changed from Algolia to OpenSearch

Your app's search position may shift as OpenSearch prioritises relevance and user satisfaction over keyword matching alone. Apps whose listings clearly explain customer value — and that have strong engagement signals — are likely to rank higher.

Ratings and reviews integrity

Atlassian takes the integrity of ratings and reviews seriously. We have the following measures in place:

• In-product attestation — reviewers must confirm they are users of the app before publishing a review.
• Programmatic enforcement — Atlassian issues written warnings to partners who falsify reviews or ratings. Failure to comply may result in suspension or removal from the Marketplace.
• Reporting — a dedicated ticket type in ECOHELP allows Atlassian team members, partners, and customers to report Marketplace violations.

FAQs

Q: Will my app's search ranking change now that OpenSearch is live?

A: It may. OpenSearch uses NLP and engagement signals rather than keyword matching alone, so rankings reflect how well your listing matches user intent and engagement. See the How search ranking works section above for guidance.

Q: How is it possible that an app with fewer installs than mine appears higher in results?

A: Many factors contribute to search rankings, and installs is only one of them. Most commonly, it means the other app's listing is better aligned to the keywords and intent behind the customer's search. See the keyword relevance and semantic match sections above.

Q: What search engine does Atlassian Marketplace use?

A: Atlassian Marketplace now uses OpenSearch — an open-source, industry-standard search platform that replaced Algolia in 2026. It supports natural language processing, custom ranking, and industry-standard quality metrics.

Q: Does Atlassian track search quality?

A: Yes. Atlassian monitors the following metrics:

• Relevance: NDCG (Normalised Discounted Cumulative Gain) and Precision
• Technical: Latency, throughput, error rate
• User engagement: Click-through rate, average click rank, bounce rate Early results show higher click rates, better conversion from search to install, and fewer "no results" searches — particularly for natural-language queries.

Q: Does geographic ranking apply to OpenSearch results?

A: No. OpenSearch ranking does not factor in a customer's geographic location.

Q: Do I need to do anything to improve my app's ranking?

A: Yes. Here's a quick checklist:

• Use clear, descriptive language in your app name, tagline, and brief description
• Write in natural language that reflects how customers describe their needs
• Avoid keyword stuffing
• Add screenshots, videos, and a demo
• Encourage customers to rate and review your app
• Provide support details on your listing For detailed guidance, see the How search ranking works section above.