As backup application vendors begin to move away from hardware, ExaGrid’s unique architecture and relentless customer focus position the company as one of the top tier backup storage players – and growing.
Something learned during the 64th edition of the IT Press Tour in New York… there’s a seismic shift happening in the backup storage market that most people aren’t paying attention to. While the giants – Dell, HPE, NetApp – continue their dominance, the backup application vendors who used to bundle storage with their software seem to be quietly getting out of the hardware business. And in that gap, a 17-year-old company called ExaGrid is quietly positioning itself as the biggest independent backup storage player in the space.
Bill Andrews, ExaGrid’s President and CEO, doesn’t mince words about the competitive landscape. “This is Pepsi Cola and Coca Cola,” he says, describing the challenge of going up against Dell and HPE. “That’s what we’re up against.” Yet despite those odds, ExaGrid has strung together 19 consecutive cash-positive quarters, maintains a 99.1% customer support renewal rate, and wins 71.3% of competitive deals when they get in front of customers.
The question is: how?
The Intelligence Engine: Market Truth From the Front Lines
Here’s something most vendors won’t tell you: they don’t really know what customers are using day-to-day. They rely on analyst reports, surveys, and partner feedback. ExaGrid does something different.
“We have 38 inside sales people that cold call all day,” Andrews explains. “16 of those are in Dublin, the other 22 are in the United States… Named accounts all day and have conversations all day. No leads. All 100% outward call quality.”
Those 38 reps, combined with 90 field sales reps spread across the globe and 58 systems engineers, are doing nothing but talking to customers about backup and backup storage. Every conversation flows into Salesforce – what backup app they’re using, what storage sits behind it, retention policies, pain points, everything.
The geographic spread is impressive: four sales reps in the UK, four in France, four in Germany, two in Saudi Arabia, two in India, two in Japan, two in Australia, four in Brazil. Andrews rattles off the list with the confidence of someone who knows exactly where his people are and what they’re hearing.
“I read a lot of stuff from people who will say, ‘oh, this is what customers think’ and I go, ‘Wow, we talked to tens of thousands a year. We’re not hearing that at all,'” he says. There’s a hint of frustration there – the kind that comes from having real data that contradicts the conventional wisdom.
The result? ExaGrid has real-time competitive intelligence that no one else can match. They know, for instance, that of all the NetBackup customers in their database, exactly what percentage are using Data Domain versus NetBackup appliances versus standard primary storage. They’ve been building this dataset for 20 years.
By the Numbers: A Company Built to Last
Let’s talk about what ExaGrid actually is before we get into why it matters. The company has been shipping tiered backup storage for over 17 years, and that singular focus is deliberate. No backup applications. No primary storage. Just backup storage.
The financial health is noteworthy in an industry where startups burn cash and established players leverage debt:
- 19 consecutive quarters of positive cash flow (Q3 2025 just completed)
- No debt of any kind – no term loans, not even an accounts receivable line
- 4,821 active customers worldwide, installed in 80+ countries
- 99.1% maintenance renewal rate – only 44 customers dropped support
- 95.2% customer retention rate
- +81 Net Promoter Score
- 71.3% competitive win rate averaged over 32 quarters
“We just finished our Q3. That was our 19th consecutive cash-positive quarter, so 19 quarters in a row we were cash positive,” Andrews notes. “Before that, we had 28 quarters in a row where we were cash break-even.”
That last stat is interesting. Twenty-eight quarters of break-even, then a shift to profitability for 19 quarters. That’s not the profile of a company chasing growth at all costs – it’s the profile of a company playing the long game.
The Largest Independent: What That Actually Means
Andrews is careful about his language when he positions ExaGrid in the market. “We are the largest independent vendor in the space,” he says, then immediately clarifies what “independent” means.
“There’s seven big storage players in the world. You all know them: Dell, Hewlett-Packard, Network Appliance, Pure Storage, Hitachi, IBM, and Huawei. Those are the seven giants. Look past that, and you look at backup storage. Nobody does more backup storage in the world than we do. I don’t care who. Nobody does more backup storage, except for those seven.”
It’s a specific claim, and he backs it up with reasoning: if you buy your primary storage from Dell, you tend to buy your backup storage from Dell. Same with HPE, NetApp, and the others. “We’re going into a NetApp shop or an HP shop saying, ‘Do something different with your backup,'” Andrews explains.
The Dell-HPE duopoly is real. And breaking through requires more than a better product – it requires a fundamental shift in how the market works.
The Market Shift: Backup Apps Abandoning Hardware
Here’s where things get interesting. A wave happened years ago where everyone wanted to sell end-to-end solutions. Commvault pushed HyperScale X. NetBackup had appliances. Everyone wanted hardware revenue on their books.
That wave is reversing – and reversing hard.
The reasoning is straightforward: software companies want Annual Recurring Revenue (ARR) from subscriptions and SaaS offerings. Storage hardware doesn’t fit that model.
If you look at subscription, why do people subscribe? One major reason: they want flexibility. They want to either do it year by year or on a three-year basis, because if something changes and they need to switch backup apps… they want to be able to switch horses.
But backup storage doesn’t change that often. The real innovation happens in what backs up the data, not what stores it. So subscription pricing doesn’t make sense for storage.
Hardware also means spare depots, field service, inventory management – none of which are core competencies for software companies. So, they’re getting out, and they’re telling customers to buy storage from whoever they want.
The takeaway?
The backup application market seems to be consolidating around four major players, all of whom are moving away from storage hardware. For ExaGrid, that’s an open door.
Technical Differentiation: Tiered Backup Storage Explained
Here’s the technical heart of what ExaGrid does differently. In the backup storage world, Andrews explained that you traditionally have two options:
Option 1: Standard disk storage (HDD or SSD). Fast backups because there’s no inline processing, but no deduplication, which means high costs for long-term retention. No backup-specific features. No air-gap security.
Option 2: Inline deduplication appliances (think Data Domain, HPE StoreOnce). You get storage savings through deduplication, but backups are slow because the system is doing compute-intensive deduplication work inline. Restores are slow because all data is deduplicated and must be “rehydrated.” And they use a scale-up architecture – you add storage shelves to a front-end controller, which means the backup window expands as data grows.
ExaGrid built a third option: Tiered Backup Storage.
How It Works
ExaGrid uses a two-tier architecture:
Tier 1: Landing Zone – This is non-deduplicated disk that acts as a high-speed cache. Backups land here at full speed, no inline processing to slow things down. The most recent backup is always stored in full, ready for fast restores. This tier is network-facing.
Tier 2: Repository – This is where things get interesting. As data hits the Landing Zone, ExaGrid starts deduplicating it into the Repository tier using what they call “Adaptive Deduplication.” It’s not inline (which would slow backups) and it’s not post-process (which would waste time after the backup completes). It happens in parallel.
Here’s the kicker: the Repository tier is non-network-facing. It’s not on the network at all. Only ExaGrid’s object code can access it. Threat actors can’t see it, can’t touch it, can’t delete it.
“The threat actor can’t see it. They cannot see it. They cannot get to it. Only our object code talks to it. It’s not connected to the network in any way, shape, or form,” Andrews explains.
Performance Claims
ExaGrid claims backups are twice as fast as Pure Storage SSD solutions. Andrews explains why: “Pure’s software is not tuned to backup. They don’t do job concurrency. They don’t pass through the backup app the exact number of streams. They’re not integrated properly with the backup applications because they don’t do that work. They don’t tweak their operating system for handling backup jobs – it’s tweaked more for transactional files.”
For restores, the Landing Zone holds the most recent full backup in non-deduplicated form. “Fast as disk restores – avoid data rehydration,” the presentation notes. With inline deduplication appliances, everything is deduplicated, so every restore requires rehydrating the data first. With ExaGrid, the most recent backup is ready to go.
For Veeam customers specifically, synthetic fulls occur on ExaGrid appliances and are “30X faster with Veeam Fast Clone,” according to the integration documentation.
Scale-Out Architecture
Every ExaGrid appliance is a complete system: CPU, memory, networking, and storage. When you add capacity, you add another complete appliance. The backup window stays fixed.
Compare that to Data Domain or StoreOnce, where you have a front-end controller and you just add storage shelves to the back. “You just keep adding disk shelves,” Andrews says. “The backup window is expanding.”
The scale-out approach also enables global deduplication across all appliances in a system, automated job management, and failure tolerance with automatic job redirect if an appliance goes down.
The Security Story: Ransomware Recovery by Design
ExaGrid’s security model is built around the reality of how ransomware attacks actually happen. Andrews walks through the typical attack pattern:
“Somebody sends out a phishing email. It looks real to your employees… I almost get caught, and I’m paranoid. I click on anything without looking at who it came from and looking at the address, right? Somebody clicks on the phishing email and it launches code.”
That code watches password usage. When IT logs into security scanners, network equipment, or backup infrastructure, the malware captures credentials. Then the attackers come in, shut down the scanners (using the stolen credentials), and encrypt or delete everything.
They have two ways they attack. One is they encrypt the primary data and then they delete the backup data. Or, they go after the backup storage first and then they encrypt the primary data.
The Non-Network-Facing Repository
Here’s where ExaGrid’s architecture becomes a defensive weapon. When attackers send delete commands – whether through CLI, through the backup application, or directly to the storage – they can wipe out the Landing Zone. But the Repository tier? Invisible.
“Our backend is not network-facing. It’s not on the network. The threat actor can’t see it.”
Andrews explained that if you’re using Data Domain or Dell PowerStore, a delete command wipes everything. The storage is network-facing and responds to those commands. With ExaGrid, the Repository holds all retention data in a tier that’s architecturally isolated.
Delayed Delete Policies
ExaGrid also implements delayed delete policies – typically 10 days, though customers can configure 5, 15, 20, or 30 days. When a delete command comes in, the Landing Zone data is marked for deletion, but the clock doesn’t start immediately on the Repository tier.
“You set it up for 10 days. Well, when this comes in and deletes on Monday, your hair’s on fire because somebody encrypted all your primary data. Just go to the ExaGrid and say we’re under attack, put it on hold. We stop the clock. We don’t delete the data.”
But here’s the problem: how do you know if it’s a ransomware attack or just an operational delete? Some companies do big data purges. How do you tell the difference?
AI-Powered Retention Time-Lock: Auto Detect & Guard
This is a new feature from ExaGrid, and it’s a genuinely clever application of pattern recognition.
The system watches deletes coming into the Landing Zone and builds patterns over time. It learns what normal operational deletes look like – how much data, how frequently, what ranges. “We’re training,” Andrews says. “AI is all about training.”
After months of learning, the system knows your baseline. Then, when something anomalous happens – say, someone tries to delete 80% of data when your biggest purge ever was 25% – the system automatically takes action.
“The first thing we do is alert IT: you could be under attack. We automatically start extending the delayed delete policy. Everything that goes by, we extend it so that it never runs out until you clear it.”
Two people – IT and Security – must log in and clear the alert before the system returns to normal delete policies. If it was a real attack, the data is protected and the clock is stopped. If it was a false positive, the dual sign-off prevents accidental data retention.
The system also alerts when deduplication ratios change dramatically – another indicator of potential data corruption or encryption.
The Go-To-Market Challenge: Breaking Through Brand Barriers
ExaGrid’s biggest challenge isn’t product capability. It’s awareness.
“If we get in front of the customer when they’re looking, it’s a rare day – like a couple percent – that we won’t get in the sales cycle,” Andrews says. “The win rate is very high. The problem for us is awareness.”
The brand barrier is real. When customers need backup storage, the first call goes to their primary storage vendor. Dell, HPE, NetApp – whoever they already have a relationship with.
So how do you compete when the incumbent owns the relationship? ExaGrid uses a multi-pronged channel strategy:
80% Reseller, 20% ExaGrid Inside Sales Cold Calling
About 80% of ExaGrid’s business flows through resellers – regional, super-regional, national VARs, and large systems integrators. The other 20% comes from inside sales cold-calling, to find the opportunity, however, sales and fulfillment are still through the reseller channel.
“We get about 20% of our business from just cold calling. People cold calling, nurturing, building a rapport over time. It takes, as crazy as it sounds, about a year to two years of building up rapport or trust with the customer before they let us in.”
The reseller channel is where ExaGrid invests heavily. Last quarter alone, the company did 220+ reseller events-kickoffs, trainings, anything to get in front of the sales forces that have existing customer relationships.
“We spend all our time continuing to build out the people who can walk us in,” Andrews explains. “If we get in front of the customer when they’re looking to get something else, it’s so rare I can’t even tell you when we won’t get in the sales cycle.”
He gives the example of Presidio, a major US reseller, walking ExaGrid into a large hotel resort chain for a $500,000 deal replacing Data Domain. Then, their park business bought another $500,000. “We would have never gotten near this customer without that reseller,” Andrews admits.
The GSI Strategy: Going After the Giants
Here’s where ExaGrid’s strategy gets ambitious. There are 14 major Global Systems Integrators – companies like HCL, TCS, Tech Mahindra, Wipro, Infosys, Cognizant, Accenture, Kyndryl, Capgemini, and others. These are massive organizations that run entire data centers for Fortune 500 companies.
ExaGrid signed HCL first. “HCL took us into lab testing. They tested us with Commvault, NetBackup, Veeam, and Rubrik… We did a resell agreement with them. We did a worldwide master license agreement with them. We have trained their nine solution centers – one in Japan, three in Europe, and five in the US.”
HCL added ExaGrid to their “Backup Next” offering. Only three vendors made that list: Dell Data Domain, ExaGrid, and VAST.
“They’re looking for two things: how do they bring their customer a better SLA, and they’ve got to meet the customer’s backup performance, restore performance, security needs, disaster recovery needs, and they can’t do it anymore with Data Domain,” Andrews explains.
Why can’t they do it with Data Domain?
“To get security with Data Domain, you need to buy CyberSense. And you need a whole other Data Domain. Double the cost. The customer couldn’t afford it. They brought us in. We have built-in retention time-lock, integrated to our product.”
More GSIs are in lab testing now. Andrews won’t say how many, but the strategy is clear: get certified by the giants who run infrastructure for the biggest companies in the world.
The MSP Opportunity
ExaGrid currently has about 100 MSP customers but has never really chased that market. That’s changing. “There’s 900 large MSPs around the world, and we’re going to start going after those this quarter,” Andrews says.
To support MSPs, ExaGrid is releasing three new features:
1. Track data by share – MSPs can export undeduplicated data usage per customer/share to bill accurately
2. Ransomware recovery by share – Recover one customer without impacting the other customers on the same system
3. AI-Powered RTL – Auto Detect & Guard
MSPs set up a share per customer on ExaGrid systems. Now they’ll have visibility into exactly how much storage each customer consumes (including all retention) and the ability to recover individual customers if one gets hit by ransomware.
The Support Differentiator: No Junior Techs, Ever
ExaGrid’s support model is unusual enough that it deserves its own section. The company assigns Level 2 and Level 3 support engineers – senior technicians – to customer groups. No Level 1 gatekeepers. No junior techs.
“If you came to our customer support meetings, you’d think it was an AARP meeting,” Andrews jokes. (AARP is the American organization for people over 55.) If you don’t know about AARP, Andrews is saying his support staff is experienced – really experienced!
“These are senior techs. They know Linux. They know networking. They know backup apps. They know storage.”
The model is simple: you get the same person every time (unless they’re on vacation). They know your environment because they installed it. They have the notes because they wrote them. You build a relationship.
The geographic spread matches the field organization: 19 support engineers in Dublin covering Europe, Middle East, and Africa, with local language support. Teams in the US, Asia-Pacific, Latin America-all in-theater, all senior-level.
“99.1% of our customers at the end of the year renew maintenance and support. 99.1%. I have only 44 customers who dropped maintenance and support,” Andrews says.
That renewal rate isn’t just about good service – it’s proof that the model works. In an industry where customers complain constantly about getting passed between junior techs and repeating their problems, ExaGrid’s approach is a competitive advantage.
Deep Backup App Integrations: More Than Just Compatible
ExaGrid doesn’t just claim compatibility with major backup applications. The company has deep, certified integrations with specific optimizations for each platform.
Veeam (9 Integrations)
- Synthetic fulls occur on ExaGrid appliances, 30X faster with Veeam Fast Clone
- Veeam Data Mover protocol support (30% faster ingest)
- SOBR (Scale-Out Backup Repository) integration for automated job management and load balancing
- ExaGrid resynthesizes Veeam Fast Clone synthetic fulls on Landing Zone for fastest restores
- S3 object store support for Veeam M365 backups
- Further deduplicates Veeam’s deduplication from 2:1 to 14:1
Commvault (6 Integrations)
- Supports Commvault compression enabled
- Supports Commvault deduplication on or off
- Further deduplicates Commvault deduplicated data up to 15:1
- Spill & Fill support for automated job management
- DASH copies for unbalanced retention
- Supports either Commvault or ExaGrid replication
Veritas NetBackup (9 Certified Integrations)
- Optimized Duplication keeps metadata and catalog current for DR restores
- Auto Image Replication (AIR) provides hot running NBU server at DR site
- Accelerator support (only changed files move)
- Granular Recovery Technology for Exchange, SharePoint, VMs
- NetBackup IT Analytics integration
- Single Target Storage Pool (automated job management, performance load balancing, global deduplication, 6PB full backup in single system)
- OST integration for faster backups and unbalanced retention
Oracle RMAN
- Multi-channel support for fastest backups
- Fast database restore in seconds to minutes from undeduplicated data
- Performance and repository load balancing
- Global deduplication across repositories
- RMAN failover if any appliance fails
- 6PB full database backup in single system
Rubrik
- Archive tier support (announced Q3 2025)
- Keeps most recent two weeks on ExaGrid plus all archives (not just long-term retention)
- 3:1 to 10:1 additional storage reduction
- Faster restores than cloud
- Additional line of defense for ransomware recovery
Scale-Out Architecture and Disaster Recovery Flexibility
ExaGrid’s scale-out model addresses one of the fundamental problems with traditional backup storage: the expanding backup window.
With Data Domain, StoreOnce, or other scale-up architectures, you have a front-end controller with fixed CPU, memory, and networking. As you add storage shelves, the controller becomes a bottleneck. Backup jobs take longer. The backup window expands.
ExaGrid takes a different approach. Every appliance is a complete system-full CPU, memory, networking, and storage. Add capacity? Add another appliance. The system automatically load balances jobs across all appliances, and the backup window stays constant.
“Every one of our appliances has memory, processor, networking, and storage. It’s a scale-out,” Andrews explains. “The backup window stays fixed-length because every one of our appliances… true scale-out.”
Multi-Site and DR Capabilities
The architecture supports complex topologies:
- Up to 16 locations in a single management group (spoke and hub)
- 50:1 WAN bandwidth efficiency across all sites
- Cross-site protection between locations
- Consolidated DR copies reduce costs
For disaster recovery, ExaGrid supports multiple scenarios:
- Second data center (customer-owned). Traditional DR setup with ExaGrid appliances in a secondary location.
- Third-party hosting/colo facility. ExaGrid appliances hosted off-site.
- Public cloud (AWS/Azure). ExaGrid has a VM that runs in Amazon and Azure. “We actually have a VM that runs in AWS and Amazon if you want to do DR to Amazon.”
- Tertiary hop. For customers worried about regional disasters, ExaGrid supports A?B?C replication or A?B and A?C topologies. “People are getting hit with floods-just turn on the news. Tsunamis, earthquakes, tornadoes, volcanoes, floods, fires… It’s nonstop. So now they want a third hop to a DR site.”
The adaptive deduplication engine deduplicates and replicates in parallel, with bandwidth throttling and WAN encryption for data replication.
The Brand Awareness Problem-and the GSI Solution
For all the technical advantages and customer satisfaction, ExaGrid still faces the fundamental challenge Andrews described at the beginning: awareness.
The solution isn’t outspending Dell and HPE on marketing. It’s finding the people who already have relationships with customers and getting them to bring ExaGrid into the conversation.
Resellers who’ve worked with a customer for 10 years and have credibility. “They walk in and say, ‘Really, look at ExaGrid. I’ve been working with you for 10 years. I’ve never let you down. I’m always here when you need me. I put half the stuff you use here in the data center.’ The customer will listen to them, which you guys might call a trusted advisor.”
Global Systems Integrators who run entire data centers. “If they’re running the data center, and they’re running the data center IT, they sometimes are an influencer. When it comes time to look at backup storage, they may influence the customer’s buying decision.”
More GSIs are in testing now. Andrews won’t specify how many, but he’s clearly playing for bigger distribution wins. “I hired the team and invested in it. It’s taken a long time to get to this point, and now it’s starting to happen.”
Where the Market is Heading
Sitting in that briefing room, listening to Andrews lay out ExaGrid’s strategy and market position, a few things become clear:
The backup storage market is disaggregating. After years of trying to bundle everything together, the backup application vendors are getting out of hardware. They want software revenue, subscription models, and ARR. Storage doesn’t fit that model, so they’re letting it go.
Data Domain appears to be on the decline. Whether Dell officially announces end-of-life or just lets it fade away, customers are noticing a lack of innovation. Two or three years without any major new features is an eternity in enterprise IT.
Security is the new battleground. Ransomware isn’t a theoretical threat – it’s a constant reality. Andrews mentions that ExaGrid customers get attacked “all the time” and recover using ExaGrid’s architecture. The non-network-facing Repository tier and AI-powered retention time-lock aren’t just features – they’re designed around how real attacks happen.
The GSI channel is the brand awareness solution. If ExaGrid can get certified and embedded in the offerings of even half the 14 major GSIs, they get distribution to the Fortune 500 without needing Dell-level brand recognition.
The Big Seven aren’t coming for this market. It’s too small for them. Dell is focused on AI servers. HPE is focused on AI infrastructure. NetApp, Pure, Hitachi – they’re all chasing bigger opportunities. Backup storage is a $6.8 billion market. That’s huge for ExaGrid. For the giants, it’s a rounding error.
The Independent’s Moment
Bill Andrews told his board 20 years ago that building what ExaGrid envisioned would take a long time. “It’s going to be highly defensible. However, you’ve got to have a big stomach because it’s going to take a long time. So you’re either in to put X amount of money into the company, or you’re not. And if you’re not, we’ll stop now.”
They stayed in. ExaGrid built the product – Tiered Backup Storage with a non-network-facing repository, adaptive deduplication, deep backup app integrations, scale-out architecture, comprehensive security, and white-glove support with assigned Level 2 engineers.
The company went through 28 quarters of break-even operations, then 19 consecutive quarters of positive cash flow. They accumulated 4,821 customers with a 99.1% support renewal rate. They maintained 71% competitive win rates over eight years.
And now, the market is shifting in their direction. The backup app vendors are abandoning hardware. Data Domain is in managed decline. Customers need ransomware recovery capabilities that inline deduplication appliances can’t provide. GSIs are looking for next-generation solutions to bring to their Fortune 500 customers.
“I’m the luckiest guy on the planet,” Andrews said. But luck favors preparation. ExaGrid built a defensible position over 17 years. Now the market is coming to them.
For VMblog readers evaluating backup storage strategies, the ExaGrid story offers a case study in how independent vendors can compete with giants: by doing one thing exceptionally well, building unassailable customer loyalty, and waiting for the market to recognize that the emperor has no clothes.
Some of the older products have a lack of new features. The backup app vendors don’t want to be in the hardware business anymore. And ransomware isn’t going away.
ExaGrid built for this moment. The question is whether they can capitalize on it before the window closes – or before one of the Big Seven wakes up and decides backup storage is worth caring about again.
Based on the numbers Andrews shared, the trajectory looks promising. Nineteen consecutive profitable quarters. Seventy-one percent win rates. Ninety-nine percent support renewal.
83% of customers chose ExaGrid when testing side-by-side with other solutions, according to the briefing notes.
That’s not luck. That’s a better product, delivered by a company that’s been playing the long game while everyone else was chasing the next big thing.
##
