Member Login

Beyond the Hype: MACH-Enabled AI for Search and Merchandising

In a few years, there will be two types of companies: those that have adopted artificial intelligence (AI) technologies and those that do not exist. Despite the hype around AI, we’re already seeing real and tangible benefits for businesses that choose to invest in it.

What’s important is that organizations carefully consider the business outcomes they are trying to achieve and plan accordingly. Every approved project should have a direct, measurable impact on costs and revenue. Using AI at organizations can actually give time back to teams to allow them to spend more time on more creative and strategic work that brings real value.

“Just about everybody can recognize that we’re in an AI hype cycle right now,” says Eli Finkelshteyn, CEO and founder of Constructor, an AI-driven product discovery platform for e-commerce. “There is a lot of pressure on companies to do something with AI. But I would advise every company to consider carefully where this technology can really add value and where and how it will drive revenue and efficiency.”

As Sree Sreedhararaj, CTO of Sephora advised in a recent webinar: “There are no cookie cutter use cases, so get specific. Look at technology adoption and use repetition. If customers reuse a service and benefit, that would be a key measure. As a retailer, we make sure it’s not just about the cool factor, but that it will change things for the customer. We’ve been at the forefront of many things and discarded them where customers haven’t been ready. E-commerce needs to remain authentic.”

This trial and error mindset is one many companies rearchitecting their tech stacks based on MACH principles are willing to embrace. And it’s important when considering where AI investments should be made. Be willing to experiment and revisit projects as AI capabilities advance.

During a session at MACH TWO, the MACH Alliance annual conference, ASICS Digital’s associate director of product management, Mindy Montgomery, said the company is re-assessing its product discovery and guided shopping experience roadmaps in light of what’s now possible with AI. MACH-certified microservices that have been built with AI embedded are strong and well tested within the search and product discovery category. It’s a good place to start a MACH journey, and prove value to the business of both AI and MACH at the same time.

Product Search and Discovery

AI-driven product discovery solutions are helping businesses to increase sales by leveraging insights from customer behavior, market trends and competitor activity. The technology is advancing so rapidly that retailers are revisiting ongoing projects that they’ve invested in to take better advantage. One of the main upsides to product search and discovery technology is that it is straightforward to implement, switch out and test ROI with – you can simply AB test it.

Let’s take a look at how MACH solutions are helping to drive adoption of AI in tangible ways.

Paul Smith

The team at Paul Smith, a global fashion retailer, installed Klevu on their existing front-end, leveraging AI search, category merchandising, and product recommendations software to boost conversion and revenue during the longer-term MACH front-end transformation project.

Adopting the MACH approach allowed the team to create a unique design, to fetch product discovery data for scalable AI content and product placement. Now, the team can serve up accurate and inspiring product and content results to shoppers faster than ever before. As a brand, Paul Smith has learned to trust the AI to deliver the right product and content results to shoppers, manually merchandising as lightly as possible, which is ideal for their lean e-commerce team.

“We wanted something that we could plug in and just leave to tick over,” said Hannah Bennett, Head of Digital at Paul Smith. “Now, we will pin key products that we feel need to be highlighted, or easily switch product ordering to favor high stock items when we are in sale mode.”


With 813 stores and 15,000 employees, INTERSPORT is the number one physical network of sports stores in France. As an innovative, customer-focused group, INTERSPORT knew that in order to improve the visibility of its product catalog digitally and improve the customer experience, they needed to increase the productivity of their team and streamline their internal processes.

Utilizing MACH certified PIM technology to harmonize product data into a single source of truth, combined with AI technology to automate the most manual, time-consuming processes like product categorization, INTERSPORT was able to improve productivity by 70%. This led to 18 hours of saved time per employee, per week. Not only were they able to scale their processing to 100,000 products a month, but they also reduced the time it takes to process new product catalogs by 50%.

“With AI, organizations can easily organize, clean, and enrich their product information, removing any inconsistencies and filling in any content gaps they encounter to ensure they’re providing the highest quality product information possible,” says Sarah Assous, Senior Director, Product Marketing at Akeneo. “Not only does this improve the customer’s experience, but it frees up employees to focus on higher-value work, such as strategic planning or customer engagement tactics.”, a leader in sustainable and innovative flooring, recently launched its B2B site with technology from Klevu and commercetools, along with Vue Storefront. The success of Tarkett is a testament to the power of MACH-certified technologies and how they can be leveraged to create unique experiences for B2B customers.

Search and merchandising play key roles in improving conversion rates for businesses targeting B2B clients. Tarkett was able to optimize the customer journey while providing personalization capabilities across all channels - from personalized search results pages to product listing pages. This enabled them to create an enriching user experience with relevant content and smooth navigation, thereby increasing conversion and creating a fast customer experience for the B2B site.

MACH’ing the Most of Your AI Investments

By 2032, the e-commerce AI market is expected to reach $45.72 billion. Eighty-four percent of e-commerce businesses place AI as their top priority, and AI for e-commerce has been shown to deliver more than a 25% improvement in customer satisfaction, revenue, or cost reduction. These are compelling statistics.

So why do these developments play so well in the MACH era? As MACH adoption increases, companies are investing in more cutting edge solutions like those delivered by our certified members. Because composable solutions can often be deployed in just weeks or months and are easier to swap out than with the old monolithic systems, you have the ability to try things. It also puts the onus on the providers to remain relevant by delivering the best technology they can, making use of the best of new technology like AI to deliver for the customer. Buyers can get the best of the best, specific to their needs.

“In the MACH community, there is inherently the mindset of being product led and customer led,” said Bettina Donmez, Digital Product Lead Cloud Platforms at PUMA during the recent MACH TWO Generative AI Learning Circle with Klevu. “MACH builds will benefit from AI technologies because MACH is the brain in itself. It connects to the different sources of truth. That MACH mindset is how I think AI will help us in implementations to solve any problem. The other thing about MACH technologies is people who are in MACH stay curious. Stay curious about the technologies they can add to their stack, or technologies they no longer need in their stack.”

“All retailers and organizations are different,” said Finkelshteyn. “MACH is technology they can play around with and try quickly, and the vendor’s job is to provide advice and iterate together. But more importantly, it’s to prove that their technology is the best technology for the use-case and drives actual ROI. That’s a big part of what MACH certification stands for. It’s technology that can be swapped in quickly, but also swapped out quickly if it doesn’t meet the customer’s needs.”