With its business rapidly growing and customer expectations rising, Thermo Fisher Scientific is turning to machine learning and robotic process automation (RPA) to transform the customer experience.
Formed from the merger of Thermo Electron and Fisher Scientific in 2006, Thermo Fisher Scientific is one of the world’s largest suppliers of scientific instruments, reagents, and services, with more than 130,000 employees worldwide. Since 2006, it has grown with additional mergers and acquisitions, including Life Technologies Corp. in 2013, Alfa Aesar in 2015, Affymetrix and FEI Co. in 2016, and BD Advanced Bioprocessing in 2018.
The rapid growth left the company highly dependent on fragmented, manual processes and disparate data sources and systems. With more than 10 million transactions and interactions per year across order entry, sales, and customer service, the company found those processes could not scale to meet the demand and deliver the experience its customers needed.
“We’re very much focused on the commercialization of acquisitions, making sure we don’t break the deal models and that things are running as they should be,” says John Stevens, vice president of IT at Thermo Fisher. “Sometimes the back-office capabilities don’t always get assimilated into the ecosystem. So, we have a lot of disparate systems across our company — ERPs, CRMs, middleware — but our go-to-market strategy for our customers, you have to make that all invisible for them.”
The problem had grown stark: Sometimes it would take customer care representatives in the Life Sciences Laboratory Products Group more than 10 minutes to answer simple calls about things like the status of an order or product availability.
“Customer care teams were going through nine different systems and nine different screens to put things together,” Stevens says. “It was an important inflection point for us as a company to ask: What are we going to do to digitally enable our colleagues inside our company to be more effective and have richer conversations with our customers?”
So, in 2020 Thermo Fisher initiated an automation and digitization project focused on eliminating those manual processes and creating a unified system of engagement for customers.
Unifying customer care
Dubbed Project Northstar, Thermo Fisher’s customer engagement initiative involved creating a series of discrete modules — order entry automation, case management, a single pane of glass search platform, and a customer insights platform — all underpinned by a centralized data lake that could consolidate relevant data from disparate platforms into a single layer.
For its order-entry automation module, Northstar leans on AI and RPA to optimize data recognition and verification, and to reduce errors and accelerate order cycle times. Although Thermo Fisher is still iterating this module, the goal is for the final release to automatically ingest fax and email orders via OCR and upload them to the system of record while routing order exceptions to a new case management portal, Stevens says.
Northstar’s case management module is an intelligent business process management platform that uses machine learning to automate tasks associated with customer emails. It auto-assigns emails to teams based on addresses or selected content and text. The module also enables customer care teams to view the entire email history and any linked emails, and to search for information related to the email within the same interface, Stevens says.
The project’s smart search platform gives Thermo Fisher employees an easy-to-use interface for pulling information from the company’s data lake on customer orders, product availability and pricing, customer-specific quotes, carrier tracking information, invoices, and more.
Project Northstar represents a shift in direction for IT at Thermo Fisher, Stevens says, one that sees Stevens and his team embracing approaches and solutions new to them.
“We didn’t have a lot of expertise in this,” Stevens says. “Five years ago, a lot of IT professionals would have been talking about putting everybody on a single ERP, a single order orchestration platform, and spending millions of dollars and months or years to consolidate that.”
For inspiration, Stevens and his team looked outside their industry to understand how other organizations have applied automation at scale. One key to Northstar’s success was Stevens’ decision to partner with professional services firm Genpact, which helped Thermo Fisher create process maps down to the click level.
“That helped us to really understand how to solve some of these very complex problems and simplify the way that our customer care teams interact with our customers,” Stevens says.
Stevens also cites the team’s focus on putting together a compelling story for company leadership as a key component of Northstar’s success.
“Not only was it going to be transformational, but it was going to inspire our colleagues to wake up every day and be more digitally enabled,” he says. “They didn’t come to work to look through spreadsheets and nine different screens and see frustration in their customer interactions.”
Of course, when IT starts talking about automation, the perception can be that IT is going to automate them out of a job, Stevens notes. As such, change management was hugely important to Northstar. Stevens and his team worked closely with the communications team and HR to ensure they were sending the right messages to end users.
Since fully deploying Northstar at the end of 2021, average call hold times have decreased 27%, and the ability to process orders faster has led to an acceleration in cash flow, Stevens says.
“There’s more selling time for the sales team because they no longer have to spend three hours a week or more calling care to ask where a product is,” he says. “We built an enterprise data lake that has all our disparate systems data so you can present it in a single pane of glass, and we use low code and no code for that, which allows us to have everything at our fingertips. We’re rolling that out now internally to our inside sales teams and to our commercial team so they can self-serve.”
Ultimately, Stevens advises IT leaders to not let themselves — and their organizations — get stuck in the technology of the past.
“You have to stay relevant; you have to stay current,” he says. “You have to convince your company that you’re going to take chances to do lighthouse programs that are going to fail fast or prove the efficacy of a program. A few years ago, we wouldn’t have thought about doing anything except an ERP migration or more integrations. Now, we’ve got these new tools that sit in front of us.”
CIO 100, Digital Transformation, Robotic Process Automation, Science Industry