small steps to
major change

Even with a career steeped in healthcare information, Josh Schoeller still learnt plenty when he was unexpectedly propelled onto the frontline. In his second week as president of Elsevier’s clinical solutions division, he took an urgent flight north to Minneapolis where his father had been taken into hospital suffering with a serious spine infection.

During the journey, Schoeller read 50 pages generated by his new team at ClinicalKey on his dad’s condition and, when he arrived, pressed for him to go straight into surgery rather than sticking with the initially prescribed course of antibiotics. “It used to be that no one questioned the doctor, right?” he says. “Today, two-thirds of every physician’s encounter is correlated to questions generated from the patient or their advocates.”

Time-pressed doctors face the challenge of dealing with concerned relatives clutching their own research as well as making sure they refer to the latest academic thinking when tackling an emergency. Complexity is everywhere. People are living longer with multiple conditions. And the total amount of medical research that was doubling about every 50 years in 1950 took only 73 days to do the same by 2020.

That explains why Elsevier, part of RELX, launched a new version of its ClinicalKey in several markets including the US in August 2021. Clinicians can still read deeply on any given condition and consult 4.8m images and more than 66,000 videos, but Schoeller’s team have also overlaid summaries digested by 150 experts for quicker reference through a cleaner, colour-coded web interface accessible on handheld devices at the point of care.

The product improvement was typical of RELX: a key element of its strategy is a deep understanding of its customers’ needs. The new ClinicalKey emerged from dozens of collaboration with users at more than 40 US hospitals and design partners from across the globe. So much talking and experimenting got to the crux of the problem. From case to case, busy clinicians have just one minute to search for information and four minutes to consume it – before they begin the hunt for solutions.

RELX does make occasional targeted acquisitions, but the main focus of the company’s strategy is to grow organically. Even in the year Covid-19 struck, the group maintained its average spend of five percent of revenues into organic investment, worth circa £350m. This steady, day-to-day improvement has helped transform RELX over the past two decades. In 2000, just 22 percent of revenues were electronic. Today, that number is 86 percent. Continuous innovation has helped the company’s shares out-perform the FTSE100 for eleven consecutive years.

Some of the organic investment used to be spent on digitising print properties. But now the focus is on leveraging analytics and machine learning, and adding extra data sets to develop sophisticated decision and analytics tools. RELX’s professional customers - from doctors and lawyers to insurers and risk professionals assessing fraud and financial crime risk - are in pursuit of better results and quicker, smarter decision.

“People like talking about disruption, but we prefer the concept of continuous innovation,” says Jelena Sevo, RELX’s chief strategy officer, who previously held various senior management roles in Elsevier and LexisNexis. She sets the tone but expects the group’s industry experts to decide where investment is needed. Tracking opportunities and pinch points involves understanding the objectives and outcomes of its customers – and RELX’s technology infrastructure means it is easy to try something out and quickly assess the results.

“We experiment every day by putting something out there, seeing whether it works, learning, iterating and then developing further.”

Schoeller, whose father is making a good recovery, is already exploring the future of clinical solutions. “Once the content is curated, our next phase of innovation is taking that content and moving it into advanced decision support,” he says.

A good example of this is Elsevier’s ClinicalPath, a “pathways” tool that guides oncologists through treatment decisions. It recommends the most effective, evidence-based pathway for specific oncology diagnosis. Recommended pathways are based on clinical input from a nationwide committee of oncologists and the latest clinical trials to help make optimal treatment decisions. A recent study found that the use of treatment pathways reduced cost and decreased hospital emergency department visits and unplanned hospital admissions in patients with stage two breast cancer. Beyond cancer, ClinicalPath may also expand into primary care.

During decades supporting the legal industry, LexisNexis, RELX’s legal business, has amassed a proud history of firsts. It was the first electronic database to transition to a web-based internet offering and in recent years has been the first to apply data analytics at scale.

This continuing reinvention has been determined not only by what the latest technology allows, but by what customers want from it. In live interviews, clients can praise or grumble about certain functions. But by tracking usage and product analytics it is easy to see where users struggle to complete a task. Such observation directs investment – and ensured the newest version of the LexisNexis’ flagship product took its cues from Google, Disney and Instagram.

“Our customers articulated a desire for a much more user-friendly, consumer-oriented product experience,” says Jeffrey Pfeifer, chief product officer for LexisNexis North America and UK. “That's a high bar in the professional space because professional results are what really matters.” His team leverages many points of customer feedback, including Net Promoter Score (NPS) verbatim comments, to gain insight into user perceptions of the product experience which are in turn used to prioritise hundreds of incremental product improvements.

Today’s competitive pressures mean the challenges are about usability and speed, as clients expect a faster turnaround of work and lawyers have a mountain of materials to review. The volume of data that LexisNexis handles doubles every year. It adds 1.7m new legal documents to its database every day, generating an astonishing 129bn new connections when the metadata has been applied that cross references each text. Crucially, the machines are not left to get on with it. Hundreds of data stewards mark the computers’ homework to ensure the artificial intelligence deployed is accurate.

Developed after feedback from more than 2,000 customers, Lexis+ was introduced in July 2020 with a cleaner design, including a simple internet search bar, and tiered pricing, common with subscription TV giants, to differentiate between the needs of small and large law firms. There is a tighter, more visual interface between traditional legal research, analytics and strategy questions.

Lexis+ also supports delivery of insights into lawyer workflow - in places like Microsoft Teams or Word - where legal workers spend hours every day. Using application programming interfaces (APIs), Lexis+ enables a user to launch a search query without having to flip screens, or switch context between product applications, to answer a client question.

RELX has around 10,000 technologists globally, but the improvements in Lexis+ were the work of 100 software engineers at the LexisNexis technology centre in Raleigh, North Carolina, as well as colleagues in Shanghai who focused on training computers to better understand an individual’s search intent. Oddly, Covid-19 helped too. The development work was largely finished when the pandemic struck, but watching lawyers work remotely in teams prompted an extra focus on making it easier to collaborate on shared document production.

Behind the scenes, none of this would have happened without LexisNexis rebuilding its technology infrastructure over the past eight years. By relocating LexisNexis’s 3.25 petabytes of data into the cloud, the company was better able to analyse customers’ pain points. The product runs at least as fast as the lawyers. LexisNexis used to be upgraded five times a year with new coding releases. Now that can take place numerous times per day.

“Simply searching for information - which has been historically a key benefit offer to customers - is no longer enough because of the massive explosion of information that is available,” says Pfeifer. “We are now building digital decision tools that enable lawyers to analyse information much more efficiently, with higher accuracy, and to ultimately, provide their clients with better legal guidance.”

For example, Lexis+ customised an open-source machine learning model called BERT (Bidirectional Encoder Representations for Transformers) developed by Google. This process is something LexisNexis refers to as 'sending BERT to law school.' This machine learning technology allows LexisNexis to offer highly relevant passages rather than presenting long lists of documents for a user to review.

What users feed into Lexis+ is evolving too. Rather than distilling their own searches from a long legal paper, they can upload that document for analysis, with relevant texts recommended back to them for reading in preparation for a case.

The mark of success is usage – and so far subscribers are spending twice as much time on Lexis+ than its predecessor. Encouraged, the team are already looking for the next tweak. “What used to take me days to do when I was a lawyer, now only takes a few minutes” says Sevo, “this allows lawyers to focus on what matters – the right legal strategy and outcome for their customers”

“It might seem to the market that Lexis+ is a one-off big product launch, but we have been adding many of these things over time, says Sevo. Adopting new technology is not easy and customers are not always ready for a completely new solution."

In common with most other people, Ash Hassib prefers not to devote too much time to searching for car insurance. “There are some companies that ask you about 60 questions before they give you their rates,” he says with a sigh. “At some point we all just give up.”

Rather than simply shopping occasionally for a quote, Hassib, senior vice president for US insurance verticals and global markets at LexisNexis Risk Solutions, part of RELX’s Risk business, is working to take the hassle out of insurance too. The volume of data collected by LexisNexis Risk Solutions means its insurer customers can provide pre-fill online forms to offer faster rates despite posing fewer questions to motorists.

The business was born from C.L.U.E. (Comprehensive Loss Underwriting Exchange), which in 1992 began collecting driver claims histories from insurers so each firm did not have to contact the previous provider before quoting and underwriting a new policy. Under the ownership of RELX since 2008, it has driven deeper into the automotive market, as well as into several adjacent insurance markets.

In 2021, 85 percent of new US auto insurance policies issued to consumers benefited from LexisNexis Risk Solutions, equating to 95 of the top 100 insurance firms. “We can't have that many more companies becoming our clients and we don't want to raise prices because that invites competition,” says Hassib.  “So, the way to drive growth is just to continue innovating.”

It helps that cars offer up more information about their drivers. Since 2012, Hassib’s team and his peers have been gathering anonymous data on speeding and hard braking from insurers’ telematics systems and usage-based insurance. And since 2018, it has added in similar details from car makers, as well as collecting specifics of driver safety systems such as parking assistance and matching it to individual cars via their vehicle identification number (VIN) so that consumers can continue investing in their safety. Underscoring its role connecting disparate industries, LexisNexis Risk Solutions has helped law enforcement bodies to automate crash reports which cuts police administration time and speeds the adjudication of insurance claims.

By knowing more, insurers can predict driver behaviour and risk and are more likely to segment drivers further, giving them the best possible price for their risk. It means good drivers don’t subsidise bad and experienced underwriters can focus their time on more complicated cases. “Once the data became more connected and more tied to an individual consumer, then the analytics for predicting likelihood of a claim really became much more powerful,” adds Hassib, who is based in Alpharetta, Georgia, outside Atlanta.

In 2010, the business decided to take its expertise into other insurance markets. First of all, it explored how its trove of auto data could be applied elsewhere, such as by using hail damage on cars to inform the understanding of rooftop hail damage for home insurers. Although not every insurer offers every product line, driving behaviour reads across to impact life premiums and home and commercial insurance can be connected if an individual runs their small business from home.

C.L.U.E. Property followed C.L.U.E. Auto, another contributory database that collects records of prior claims and has accelerated processes. Applicants need only fill in five questions to receive a quote, down from 40. And thanks to the work of data scientists in the UK, Ireland and China, insurers can be advised which homes need inspections as a matter of priority.

But the biggest transformation has come in life insurance, where data use had been slow to catch on. Now, by syncing medical information with credit histories and public records, what took 45 days including a thorough medical examination has reduced so that insurers can offer a policy within minutes or hours of application depending on the insurer's process and without a visit from the nurse.

From its beginnings in automotive, some 65 percent of the business’s growth today comes from innovative new products that did not exist seven years ago – much of it in new markets.

“We employ many professionals that worked in the industries we serve, and we take a very long view on the market. In the auto insurance market, we may not know exactly how long the transition to connected or driverless cars will take, but we know it will happen. Consequently, we are already building datasets and analytics that will support risk decision making in the future.”  

The atmosphere at drinks industry gathering Imbibe Live was understandably convivial. Held over two days in September 2021, the event was an opportunity for more than 5,000 visitors to gather for tastings, demonstrations and discussions about the latest beers, wines, spirits and soft drinks. And, after an 18-month Covid-19 enforced restrictions, they could finally do so in person, at London’s Olympia destination.

Among the wholesalers, caterers and cocktail bar owners trying low-alcohol lager and high-value Japanese sake, Gaby Appleton was encouraging delegates to sample another new launch. Emperia, an app designed by RX, RELX’s recently rebranded Reed Exhibitions, collects and segments sales leads by scanning an individual’s QR code. It was used by 98 percent of Imbibe attendees.

“The purpose of most trade shows is engineered serendipity: expanding horizons, finding markets, customers or suppliers you didn't already know,” explains Appleton, RX’s new chief digital product officer who started her career at RELX with Elsevier in 2010 and spent her lockdown seconded to the UK National Health Service developing the NHS Covid-19 digital contact tracing app. “Emperia makes the whole process of capturing your leads smoother and means a better return on investment.”

The service was under consideration three years ago, long before shutdown forced event organisers to think exclusively digitally. Executives in RX’s Australian division were pondering how they could come up with a single leads product rather than relying on niche registration partners that differed between events and upsold their own apps to delegates.

Emperia was trialled at a handful of US shows in spring 2019, including at JCK Las Vegas, a vast annual event for the jewellery trade held at the resort city’s Sands Expo. The gradual rollout meant developers had early warning of hiccups that occurred when RX was running two shows in the same venue.

Appleton’s plan is to capitalise on RX’s scale and technology capability in other areas. For consistency, delegate feedback from all its 400 events in 43 industry sectors is now all processed through one centre in the Philippines. And improvements do not wait a year to be implemented when the same show rolls around again; if applicable they inform how the next show in the network is put on.

Unsurprisingly, Covid-19 led to a raft of digital innovation in lieu of physical events. “Now we're working through which of those really worked and are worth doubling down on and what didn't work,” says Appleton. “I try to coach my team to focus on what is the problem you are solving for the customer, not what feature do you want to build.” That involves using digital engagement tools to reach a wider audience and capture more data about their interests, complementing the returning face-to-face events."

The leading travel show, World Travel Market London, staged a virtual version five days after the in-person event at London’s ExCeL centre in November 2021, so exhibitors could host more one-on-one meetings for people who did not travel. Science fiction fans were not so patient, snapping up digital tickets for the sold-out New York Comic Con that offered on-demand content so they could watch from afar appearances by Stranger Things’ David Harbour and Star Trek’s George Takei. The thrice-yearly Functional Fabric Fair has gone one better, launching Performance Days Loop, a year-round portal that offers product recommendations and professional matchmaking for material makers and clothing designers.

Appleton is an example of RELX’s strategy of cross-fertilising leaders across its divisions to share ideas and leverage expertise and skills across the organisation. Sevo points to the sharing of technology, data and capabilities across the group as well as forums led by the chief technology officers and chief product officers that meet every other month to discuss group-wide best practice. There is also an intranet through which anyone can seek advice from RELX colleagues that have experimented with the latest artificial intelligence and data science approaches.

Back at Elsevier’s Clinical Solutions, Josh Schoeller is another senior internal transfer. But rather than leaving behind LexisNexis Risk Solutions, where he was chief executive of the healthcare business, he is tasked with finding gains from closer working between digital decision tools that help clinicians, and healthcare insurers, providers and life science companies with theirs.

Read more about RELX's acquisition strategy in this story.