Article | 5 min read

What are data silos and why they’re a problem (+ how to fix it)

Data silos make life harder for support agents and customers alike. Learn how to free siloed data and improve CX.

By Patrick Grieve, Contributing Writer

Published June 10, 2021
Last updated December 16, 2021

It’s impossible to provide great customer service without great customer data. You can’t give people the white-glove treatment if you don’t know anything about them.

Especially in today’s data-driven world, companies need to leverage their customer data to make informed decisions and improve their business. Yet some organizations struggle with outdated or insufficient information. Other brands capture plenty of data but fail to connect it all because it lives in multiple sources and systems. Both situations can lead to the creation of data silos.

Data silos cause inefficiencies and complicate customer service interactions, leading to stressed-out support agents and unhappy clients. Sales, product, and marketing teams could also be missing out on valuable insight about your customers. That’s why it’s important for businesses to understand how information silos are built and how to eliminate them.

What are data silos?

A data silo is a set of information that can be accessed only by certain groups within an organization.

Saying that information is “siloed” is another way of saying that it’s been put somewhere out of reach.

“A data silo occurs when you have information that pertains to a customer or a business and it’s stored in various places,” says Jessica Mills, a senior product and platform manager at Zendesk. “That’s not accessible for a lot of teams—or not accessible at all.”

In a customer service context, that means support agents might not have all of the information they need when interacting with a customer. This can hinder agents from providing a positive experience.

Why do data silos occur?

Unfortunately, data silos can easily occur. Many organizations—especially large ones—naturally tend towards siloing. But why does this separation of data occur?

According to Neal Mhaskar, Zendesk’s senior technical product marketing manager, there are two factors that lead to information silos, and they’re often intertwined.

  • Technology: Organizations often rely on multiple business tools and different databases, which may not be fully integrated with one another. This can lead to poor data management. “These systems don’t necessarily talk to each other right out of the box,” Mhaskar explains. “So, there could be differences in protocols or functionality, which result in the inability for one system to talk to other systems—for instance, if they’re older systems that a company has built itself.” In other words, relevant data often lives in different sources that aren’t connected on the back end, which makes data sharing difficult.
  • Company structure: At the same time, there’s often a mix of organizational issues that exacerbate the issue. “Different departments within an organization can be siloed, and their data can be siloed in the same way,” Mhaskar says. For example, a company’s sales team may not collaborate with the support team. As a result, sales agents can’t look at customer service interactions to identify potential upsell opportunities or to determine if an account has an open issue. “I would say it’s a mix of technical deficiencies and a lack of organizational impetus that keeps companies from breaking down silos,” says Mhaskar. It’s important to build a company culture where collaboration is not only encouraged, but teams also have the tools and technology they need to make it happen.

Why are data silos bad for business?

Obviously, organizational silos aren’t ideal. But beyond vague notions of reducing collaboration and transparency, there are three major reasons why data silos are bad for organizations.

1. Reduced agent productivity

“Specifically within customer support, data silos force your agents to spend a lot of time trying to track down the right information,” explains Mills. “They’re sometimes jumping between four to 10 to 20 systems just to troubleshoot a problem.”

Your support team could save lots of time and effort if they were able to easily surface the information they need. Productivity also increases when agents can easily collaborate and share knowledge across the company.

2. Increased costs and inefficiency

Decreased productivity can also have a big impact on your bottom line.

“When you don’t connect your systems, sometimes it also costs you more money.”Jessica Mills, senior product and platform manager at Zendesk

“The inefficiencies add up pretty dramatically,” Mhaskar warns. “For a company with a hundred agents, each agent spending just five percent more time over the course of the year is pretty substantial. That’s easily an entire head count or multiple head counts, in terms of time spent on these issues.”

That’s a pretty conservative estimate, too. Mhaskar believes data silos can easily increase the handle time for an agent by 10 or 20 percent.

“When you don’t connect your systems, sometimes it also costs you more money,” Mills adds. “You have all of these systems, and your agents need access to all of them. So, you have all of these unnecessary licenses, too.”

3. Poor customer experience

Aside from costing your business time and money, data silos also create a burden that’s ultimately the customer’s to bear.

“Reduced agent productivity and lack of access to the right information ends up impacting customer satisfaction,” Mills says. “It can create negative experiences where agents are asking customers for information that they should already have or customers are waiting for the agent to track down the right details.”

Data silos are often responsible for the worst types of customer service interactions: long wait times, endless transfers, and customers being forced to repeat themselves over and over again.

But when businesses break down information silos, agents can access the data they need to quickly resolve a customer’s issue.

4 ways to break down data silos

Knocking down data silos often requires new technology and initiatives. Businesses should decide what type of customer experience they want to provide, then empower their support agents to deliver that kind of care.

1. Map out the ideal customer journey

“At an organizational level, I think there’s some cross-functional collaboration that can be done,” says Mhaskar. “Lots of organizations will create a task force to work on issues around unifying customer data.”

From a CX perspective, breaking down silos starts with identifying objectives. Begin by determining the type of experience you want to provide customers and what data is required to create it.

Some businesses use customer journey mapping to chart every interaction a customer could have with their brand. Then, they analyze what sort of information they need to optimize each stage of that journey.

“It’s not just a concrete set of steps, though, because there are more qualitative aspects of that journey as well,” Mhaskar explains. “Are you trying to make this journey personalized? Are you trying to create unique, special journeys for customers as they interact with your organization?”

Not every company will share the same goals, so it’s important to first establish your organization’s priorities.

2. Find out what your agents need to know

Your support agents are obviously going to play a huge role in crafting customer experiences, so it’s important to determine what data they need to fulfill your CX objectives.

“It’s worth taking a holistic look at the tickets that agents deal with and determining the top five or 10 most common issues,” Mhaskar recommends. “Then ask: What would help agents solve these issues more quickly? What could help them reduce the handle time or be more efficient? What tools are agents toggling between?”

While the goal is to break down data silos, make sure you don’t overcorrect and overwhelm agents with too much information.

“It’s not efficient to give agents access to every piece of data that they might need across different silos,” Mhaskar says. “But if you look at the most common issues agents deal with, that might address 80 percent of tickets. So, what type of data is required to help solve those issues faster?”

“Lots of organizations will create a task force to work on issues around unifying customer data.”Neal Mhaskar, senior technical product marketing manager at Zendesk

3. House all your data in one place

From a technical perspective, the best way to avoid information silos is to place your customer data coming from multiple systems under a single source of truth. Instead of storing the data across different software, connect it all through an open, flexible CX platform like Zendesk Sunshine.

“Sunshine gives you a place to unify your data to create a single view of the customer,” Mhaskar explains. “It can be used to give agents more context and to personalize and customize experiences for customers.”

As a holistic platform native to Amazon Web Services, Sunshine can also eliminate data silos that occur when businesses have multiple incompatible tools.

“It can often be confusing as to where and how some of this data should be unified,” says Mhaskar. “Historically, there hasn’t been a single system for agents to use that has all of this data. Other organizations often own these systems, so there could be a sales CRM and there could be separate databases associated with ecommerce activities.”

Sunshine lets you connect and understand all your data—wherever it may live—and use it across your business, giving you a central place to create that single view of a customer.

4. Give agents the context they need

You understand the type of information your agents want access to, and you have that data stored in one place that’s easily accessible—now what?

You need to provide your support team with that important info in a way that doesn’t overwhelm them. For example, Zendesk gives agents an “essentials card” that includes all the relevant details about the customer they’re engaging with. The card may contain basic information (like the customer’s name and preferred language) or more in-depth data (such as customer status or membership level).