Back to Customers

How Pick n Pay Drove More Relevant Search and 8X ROI on Incremental Revenue

8X
ROI on incremental revenue
Industry
Retail & Grocery
Headquarters
Cape Town, South Africa
Solutions
Search, Recommendations, Searchandising, App Experiences
8X
ROI on incremental revenue
Go to the Results

Summary

Executive Summary

Pick n Pay, one of South Africa’s largest retailers, needed site search that could keep pace with a fast-growing ecommerce business. As online demand increased, its legacy search experience often returned poor matches, created friction for shoppers, and put extra pressure on a lean ecommerce team. 

Pick n Pay chose Constructor to bring more accurate, intent-based search to its digital experience and build a stronger foundation for future growth. Since launch, the retailer has attributed more than 8X ROI on incremental revenue to Constructor and expanded the partnership from website search to recommendations, searchandising, and app experiences — accelerating omnichannel growth.

About Pick n Pay

Pick n Pay is one of South Africa’s largest and most trusted retailers, serving millions of shoppers across grocery, household essentials, clothing, and more.

As ecommerce adoption grew, Pick n Pay made strategic investments to modernize its digital experience and deliver a fast, intuitive, and reliable online shopping journey.

Go to the Results
PickNPlay-about@2x
PickNPlay-about@2x

As ecommerce adoption grew, Pick n Pay made strategic investments to modernize its digital experience and deliver a fast, intuitive, and reliable online shopping journey.

Go to the Results

The Challenge

Poor Search Results Slowed Shoppers Down

When Pick n Pay’s ecommerce business began scaling, its legacy search solution could not keep up with ranking its wide range of products.

Day over day, customers struggled to find what they needed.

Elizabeth Bishop, Head of Ecommerce, summarized the frustration plainly:

“Every time I searched, the results were completely inappropriate.”

PickNPlay-challenge-1@2x

When Pick n Pay’s ecommerce business began scaling, its legacy search solution could not keep up with ranking its wide range of products.

Day over day, customers struggled to find what they needed.

Elizabeth Bishop, Head of Ecommerce, summarized the frustration plainly:

“Every time I searched, the results were completely inappropriate.”

PickNPlay-challenge-1@2x

The team saw several issues again and again:

  • Poor relevance for common queries: Searches like “red wine” surfaced hair dye called ‘red wine’ instead of actual red wine. And top-selling household essentials brands, like the market-leading dishwashing liquid, were buried in search results. (This was especially problematic during peak, end-of-year shopping periods, such as Black Friday. Around this time, South Africans use their bonus salaries to stock up on household essentials and prepare for Christmas.)
  • No intent understanding: The system relied entirely on keyword matching, leading to mismatches, missed conversions, and shopper friction (e.g., hair dye being returned for the query “red wine”)
  • Limited internal bandwidth: Because Pick n Pay’s ecommerce team was small, with minimal merchandising support, fixes were sporadic and reactive
  • Risk of breaking other queries: Fixing one search term often created new problems in others

The company outgrew its legacy provider, Solr, and never viewed building search in-house as the right answer. Pick n Pay builds many of its own systems, but leadership saw search as a highly specialized problem better handled by experts.

As Enrico Ferigolli, Retail Executive - Omnichannel, put it:

“To build search yourself, you need very specialized expertise. There are many things that make sense to build internally — this is not one of them.” 

The Solution

Why Pick n Pay Chose Constructor

Pick n Pay evaluated several platforms, including Algolia. While alternatives offered appealing UI and “bells and whistles,” they still relied on the same keyword-based approach that had already failed to scale.

PickNPlay-wide@2x

Constructor stood out for three reasons:

  • Intent-based search and personalization

    Pick n Pay wanted a solution that could do more than match text. The team needed search that could better predict what shoppers were actually trying to buy, especially in a grocery setting where speed and accuracy matter
  • White-glove partnership

    The team also valued the way Constructor worked with them from evaluation through launch. Elizabeth described the level of engagement as “the gold standard.” This included hands-on support, quick implementation, and a relationship that stayed close well after the first launch
  • A strategic growth roadmap

    Pick n Pay wanted to expand into recommendations, searchandising, and long-term AI-driven experiences. Constructor’s roadmap aligned directly with these ambitions

Implementation

How the Initial Rollout Led to Wider Expansion

Pick n Pay launched Constructor on the website first, then extended it to the shopping app through a separate index and integration. Later, the retailer added search to a relaunched loyalty-led app experience and introduced another index to support clothing.

That rollout did not stop at a single deployment. Pick n Pay and Constructor continued building together across multiple surfaces and use cases, turning the initial search launch into a broader product discovery program.

PickNPlay-implementation@2x

Pick n Pay launched Constructor on the website first, then extended it to the shopping app through a separate index and integration. Later, the retailer added search to a relaunched loyalty-led app experience and introduced another index to support clothing.

That rollout did not stop at a single deployment. Pick n Pay and Constructor continued building together across multiple surfaces and use cases, turning the initial search launch into a broader product discovery program.

PickNPlay-implementation@2x

The Results

Stronger Search, Smoother Shopping, and Measurable ROI

Since implementation, Pick n Pay has seen:

8X
ROI on incremental revenue
+5.76%
in checkout revenue
+7.96%
in purchased items per user
+3.12%
in add-to-carts

Internally, the reaction was immediate. Elizabeth described the change as “night and day.”

Search began working more like shoppers expected it to work: faster, more accurate, and more consistent across categories. Customers who once struggled with broad, practical shopping missions (such as weekly replenishment or household restocking) now receive seamless, accurate results. The need for emergency synonym fixes and manual search tuning dropped away. 

A True Partner in Innovation

For Pick n Pay, the value has gone beyond search performance alone. Pick n Pay emphasizes Constructor’s collaborative approach as a core value driver, a vendor that’s engaged, thoughtful, and willing to challenge assumptions and bring forward new ideas.

That mattered because Pick n Pay was not looking for a point solution. It wanted a partner that could help the business keep improving digital shopping over time, across channels and use cases.

Enrico’s view of the road ahead captures that well:

 

“The future is bright. We want to implement a lot of features and make shopping really easy.”

Looking Ahead

What's Next

Pick n Pay continues to expand its use of Constructor’s platform, with several priorities underway:

  • Rolling out additional search and discovery capabilities across more digital surfaces

  • Expanding recommendation touchpoints

  • Deepening searchandising strategies to support loyalty, promotions, and seasonal moments

  • Exploring AI shopping agents to make complex grocery missions effortless

  • Continuing to optimize apparel search with a dedicated index

With a strong foundation and a shared commitment to test-and-learn optimization, the partnership is positioned for continued revenue and experience gains.

PickNPlay-proof-schedule@2x
PickNPlay-proof-schedule@2x

Pick n Pay continues to expand its use of Constructor’s platform, with several priorities underway:

  • Rolling out additional search and discovery capabilities across more digital surfaces

  • Expanding recommendation touchpoints

  • Deepening searchandising strategies to support loyalty, promotions, and seasonal moments

  • Exploring AI shopping agents to make complex grocery missions effortless

  • Continuing to optimize apparel search with a dedicated index

With a strong foundation and a shared commitment to test-and-learn optimization, the partnership is positioned for continued revenue and experience gains.

Powered By:

See how leading retailers drive measurable, sustained growth with Constructor.

Nearly 70% of shoppers say retail site search still needs an upgrade, but executives don’t just want better search — they want compounding results that move the P&L quarter after quarter. Constructor’s AI-native platform reasons over shopper behavior, context, and product data to surface the right products at the right time — and it keeps learning to drive incremental lift long after launch. Request a demo to see how we partner with enterprise retailers to deliver provable ROI at scale.

Request a Demo
  • Understand the revenue impact of AI-powered search and discovery
  • Explore continuous optimization to lift RPV, AOV, and conversions over time
  • Learn how AI reasoning interprets shopper behavior and intent across every touchpoint
  • Learn how our Proof Schedule de-risks your investment with real data

Nearly 70% of shoppers say retail site search still needs an upgrade, but executives don’t just want better search — they want compounding results that move the P&L quarter after quarter. Constructor’s AI-native platform reasons over shopper behavior, context, and product data to surface the right products at the right time — and it keeps learning to drive incremental lift long after launch. Request a demo to see how we partner with enterprise retailers to deliver provable ROI at scale.

Request a Demo