Intelligent search powers ecommerce websites, site search engines, portals to support hybrid working, and customer support workflows. We’re going from the early days of keyword search boxes to today’s natural language query interfaces. The intelligence will progress from today’s automatic relevance tuning machine learning algorithms onto tomorrow’s pre-trained models and program synthesis.

This was the nature of my conversation with Paul Nelson, managing partner at Accenture, and Ciro Greco, vice president of  AI at Coveo, during a recent webinar covering insights on intelligent search in driving digital transformation.

During the webinar, we covered several essential questions around intelligent search, how progressive organizations use them to compete for customers, and how evolving algorithms improve end-user experiences. 

The Future of Intelligent Search in Customer Experiences

Should you upgrade the website or improve the ecommerce store and leave the search experience to basic keyword searching and category-based navigation? Greco believes that customer experience is more than an assembly of technologies. He says, “I like to think of a website as an organism, so you don’t want to optimize one organ and not the other, and they have to work in harmony.”

Nelson provides some insights into what it means for a customer-facing website to “work in harmony.” He focuses on the external information and data sources that can enrich the experience. “Intelligent search is not just for ecommerce – but for everything and is about including information from anywhere you can get it,” says Nelson.

He then lists a mix of internal data sources such as customer support platforms and external sources of product information that provide context and richer experiences during the buying process.

My point of view: Experiences need to achieve loyal buyers on ecommerce sites, frequent readers in media, and enthusiasts for all brands. Product managers can think of intelligent search experiences as bringing external information to their users, which will drive key user behaviors like growing repeat visits and increasing the number of buyers.

Hybrid Work Is the Future of Employee Experiences

One of my questions for these experts was to better understand a key finding from recent research. The research shows 85 percent of companies increased their investment in enterprise search over the last 12 months, with almost 50 percent labeling it significant or dramatic increases.

Nelson shared his insights on why intelligent search is a critical capability to support hybrid work. “People are dealing with the digital view of the organization and discovering that search is a critical component to get people to the content and the knowledge. Because you can’t just bump into somebody in that hallway and for someone to introduce you to the right expert.”

Greco also reminds everyone why enterprise search is challenging to implement. “There’s a lot less behavioral data to infer context, so you have to capitalize on the knowledge encoded from many company data sources.”

My point of view: With more organizations committing to hybrid working and trying to derisk the impacts of the great resignation, IT leaders must find ways to deliver enterprise search experiences without the tech complexities. IT leaders should focus on loading, cleansing, and enriching data sources so that an intelligent search platform can rank the results and optimize the experience. IT can accelerate developing department-specific experiences and integrations with low-code search capabilities and use headless search architectures when implementing more advanced experiences.  

Want Intelligent Search? Focus on the Data!

Should you focus on the machine learning models when upgrading search to support the digital workforce, improve customer experiences, or enhance customer support capabilities?

Nelson has a very simple recommendation that should be the basis of search implementations. He asks, “What is the goal of search? “ and then answers his question with, “It’s to reduce the distance between the user and the information that they need.”

Greco follows with a simple recommendation. “Put the data in one place,” implying that IT leaders should focus on integrating, cleansing, and joining data sources and optimally structuring it for search. He goes on to say,  “Not understanding small things about data management and data engineering torpedoes a whole bunch of stuff that you do later on.”

So even though a lot of today’s media and interest is around the algorithms, Greco believes most leaders should invest their efforts in aggregating data and improving data quality. Greco says, “The marginal gain that an enterprise has adopting machine learning-based applications is spending time and resources in making the data good. Capitalizing on whatever downstream application you want to build on top of this data, being that BI or machine learning model, is going to be way so much better with cleaner and enriched data.” 

And Nelson warns listeners around pursuing what may look like an easy path. “Installing an open source search engine and making it work for you, that’s a long road,” Nelson states.

My point of view: The insight here is that intelligent search requires product managers, UX designers, and technology architects to focus on the inputs (the data!) and the outputs (the experience!). In other words, avoid a do-it-yourself to the models and algorithms because these can be bought, managed, and upgraded with an intelligent search and recommendation system. The future of search is in pre-trained models developed with billions of parameters, but the business team must own the data and manage the experience.

Watch the webinar to learn more insights on the present and future of intelligent search in driving digital transformation.

This post is brought to you by Coveo.

The views and opinions expressed herein are those of the author and do not necessarily represent the views and opinions of Coveo.