Comparing Algolia vs Athos Commerce vs ExpertRec is more often than not, a consumer doesn’t search an online shop in the way the words in the product catalogue are written.
For example, a person might simply say “lightweight laptop for coding under ₹70,000”, while the product collection may be full of products that can be classified in terms of processor, RAM, weight, display, and operating system. This is where the search engine has its task, which is much more complex than searching for keywords.
Ecommerce search evolves from keyword-based search to semantic retrieval, hybrid-search, natural language processing, personalization, and artificial intelligence-based ranking.
Algolia, Athos Commerce, and ExpertRec solve this problem in their own ways with different product solutions.
The search problem is no longer just about keywords
Traditional ecommerce search involves searching through an indexed catalog and comparing query terms with fields.
Modern systems include additional steps:
Query marine processing → intent determination → candidate tracking → filtering → ranking and selection and merchandising → delivering results.
These differences become crucial in long-tail searches.
Let us examine the following phrase:
“waterproof trekking backpack with laptop compartment below ₹5,000.”
- Waterproof – quality
- Trekking backpack – category and application
- Laptop compartment – quality
- 5,000 INR – price limit
The number of supplementary signals in the catalog drive the importance of the search engine’s capacity to correlate natural language searches and product data.
Algolia: an API-first search architecture
Research shows that search-computers usually have been found to spend 2.6 times more than non-search computers, while AI-powered search is increasingly being regarded as part of the ecosystem of e-commerce technology, according to Algolia.
Algolia prides itself on being an API based search and discovery leader. The current AI Search solution takes advantage of combining keyword and semantic methods and offers hybrid search capabilities, AI ranking, personalization and query categories, along with autocomplete, data analytics and merchandising.
The straightforward implementation may look like the following:
Product catalogue→indexing pipeline→Algolia index→search API→website.
This framework allows developers a lot of freedom in terms of how they want their search to be integrated into their solutions.
The technical challenge is that the search process does not finish with the indexing. The teams may also be responsible for:
- searchable attributes
- ranking configuration
- filters and facets
- synonyms
- query rules
- autocomplete
- analytics
- personalization
- merchandising
- index replicas
- frontend integration
Algolia claims to have more than 18000 clients in over 150 countries and says that it performs more than 1.75 trillion searches in a year. Also, the engine maintains more than a billion queries each five seconds and provides average speed of response lower than 20 milliseconds.
Athos Commerce: search connected to merchandising
Athos Commerce is focused on an advanced product-discovery. Today, the platform combines the tools of search and personalization, merchandising and product feed management, as well as the recent AI and agent capabilities. Athos claims that its search technologies can recognize the intention of buyers rather than only strictly referring to keywords; while the merchandising tools enable e-commerce marketers to manage the location of items, marketing promotions, and changing positions of products in the ranking process.
Suppose a clothing chain is flooded with a search like:
“linen trousers for summer under ₹3,500”
The approach leads to the establishment of a different operational model.
In this case, the retrieval system should find out whether a specific item matters, while the merchandising department may take advantage of its capabilities and:
- encourage purchasing from the assortment;
- place priority on a seasonal collection;
- push certain products;
- move out of the ranking goods that have low stock level;
- Experiment with the ranking strategy.
Athos is firm in its support of merchandising via search and category experiences with activities such as promotion rules and A/B testing. The company reports that its platform is being used by more than 2,700 brands around the world. It has also reported user-specific results including a 300% jump in search-induced revenue at John Smedley and a 951% rise in revenue per visit at Michael Stars. The numbers stated in this paragraph represent case-study results from Athos and should not be taken as general expectation.
Where ExpertRec fits
ExpertRec takes the approach of looking at the issue from the perspective of onsite search, so it has integrations that can be used on top of the current e-commerce technology stack. Its current functionalities include semantic product search with completion, spelling correction capabilities, advanced filtering capabilities, search ranking, synonym, personalization, and analytics.
Also, it has a solution for SKU and metafield search, variants, make-and-sell capabilities, and recommendations.
This is important when products are defined in multiple fields.
For example, the request like: “cotton kurta for wedding guests below ₹2,800”should not be perceived as one string.
The search agent may interpret the following:
- cotton → material
- kurta → type of goods
- wedding guests → contextual intention
- ₹2,800 → price condition
Accordingly, this defines the retrieval process as a combination of text relevance, product characteristics, filtering, and ranking.
Additionally, ExpertRec can allow sellers to specify searchable product characteristics and implement controls of ranking characteristics, filtering, synonyms, and merchandising.
AI-powered ecommerce search and product discovery
Connect shopper intent with the right products using ExpertRec
ExpertRec combines semantic search, autocomplete, typo correction, filtering, ranking, personalization, SKU and metafield search, and search analytics to support ecommerce product discovery.
Semantic search • SKU & metafield search • Analytics
Exact SKU queries need a different treatment
Natural language discovery forms only a part of the ecom search scenario.
Take for instance the case of
“SKU-QT5816”
In such a case, it can be confidently assumed that the customer is not expecting the search engine to find broadly similar items. What they want is their SKU to appear in the search results.
This is an exact match retrieval problem.
The Shopify search provided by ExpertRec also specifically talks about SKU search, metafield search and variant display.
It is important to note this difference in case of stores selling electronics, industrial components, spare parts, items of fashion, etc.
Search analytics turns queries into product intelligence
Search analytics can identify problems that catalogue analytics are unable to do.
For example, when 12,000 people search for products in a month one shouldn’t just look at the number of searches.
Some of the metrics to investigate are:
- Zero results
- Click-through rate
- Searches without clicks
- Top queries
- Queries which are going up
- Spelling mistakes
- Search to purchase behaviour
- Revenue from search
- Queries that need adjustments
ExpertRec’s reports include such metrics as search-volume change, click-through rate, number of searches without results, unclicked queries, etc. In addition, the company has a feature which connects searches with revenue from products and categories.
Athos offers reporting on searches as well, which includes query metrics and links for merchandising purposes.
Algolia also offers search analytics and merchandising as well as tools to test ranking methods in order to improve search performance.
Algolia vs Athos Commerce vs ExpertRec
| Capability |
Algolia
|
Athos Commerce
|
ExpertRec
|
|---|---|---|---|
| Core approach | API-first search and retrieval | Commerce-wide product discovery | Ecommerce search and discovery |
| Semantic search | Yes | Yes | Yes |
| Hybrid search | Yes | Yes | Available through AI search capabilities |
| Autocomplete | Yes | Yes | Yes |
| Typo correction | Yes | Yes | Yes |
| Faceted filtering | Yes | Yes | Yes |
| Search ranking controls | Yes | Yes | Yes |
| Merchandising | Yes | Strong merchandising layer | Yes |
| Personalization | Yes | Yes | Yes |
| SKU search | Supported | Supported | Supported |
| Search analytics | Yes | Yes | Yes |
| GA4 integration | Available through integrations | Commerce analytics ecosystem | Yes |
| Shopify integration | Yes | Yes | Yes |
| Product-feed management | Integrations/data tools | Core platform capability | Integrations |
| API/headless integration | Yes | Yes | Yes |
| Primary technical consideration | Search infrastructure and developer control | Unified discovery and merchandising | Search implementation, relevance and ecommerce controls |
The architecture matters more than the feature checklist
While it is true that the three platforms have a lot of features in common and it is tempting to compare them feature to feature, this would be quite misleading.
The better questions to ask are related to architecture, i.e.:
- How are the user queries translated into the value that can be searched?
- Can product identifiers, such as SKU, eliminate semantic ambiguities?
- What business rules are applied after the retrieval process?
- Can the ranking of products be changed without involvement of engineers?
- Will it be possible to relate the results of search to conversion and revenue?
The answers to the above questions show the difference between just a search box that shows products and a search system that becomes an integral part of ecommerce industry decision-making process.
In case you are already operating on Shopify, Magento, BigCommerce or any other ecommerce platform, ExpertRec lets you build a search layer without changing your existing ecommerce platform. The company claims to support over 100 types of content through connectors and offers an easy API for integration with existing systems.
The takeaway
Ecommerce search has progressed from word matching to understanding the shopper’s intent.
The API-first approach of Algolia has advanced its relevance and analytics capabilities. Meanwhile, Athos Commerce has integrated search into its personalisation, merchandising and product feed management. For its part, ExpertRec has combined AI search, semantic search, autocomplete and filtering, ranking, analytics and e-commerce-specific discovery capabilities.
When evaluating these platforms, what is important for an e-commerce team is not which platform has the most features but whether it can understand the language used by shoppers, connect this language with the product catalogue, and convert this interaction into real products discovery.
