Netduma CEO & Co-founder Luke Barlow says the best place to understand and optimise the broadband experience may not be in the cloud at all. Instead, Netduma puts intelligence directly on the home gateway to continuously measure what is happening, identify the root cause of problems, and in many cases fix issues locally and automatically. Here’s a summary of WiFi NOW’s recent Tech Watch interview with Netduma.
For more than a decade, broadband operators have been moving network intelligence, analytics, and customer support functions into the cloud. Netduma CEO & Co-founder Luke Barlow believes that when it comes to understanding the actual broadband experience inside the home, the industry may have taken this idea too far.
“The experience happens in the home. So why wouldn’t the intelligence?” is how Netduma sums up its proposition. In the latest edition of TECH WATCH, Luke Barlow explains why Netduma has taken a fundamentally different approach to home connectivity optimisation. Rather than periodically collecting information from a broadband gateway and analysing it remotely, Netduma runs its intelligence locally on the gateway itself. That means monitoring the home connection continuously, understanding which devices and applications actually matter, identifying the root causes of poor experiences, and potentially taking corrective action while the problem is happening.
From fixing gaming lag to optimising the whole home
Netduma’s origins go back to 2014 and a very specific problem: Online gaming lag. Luke Barlow says the company’s founders were gamers themselves and wanted to understand why an otherwise good broadband connection could suddenly produce a terrible gaming experience. That work eventually led Netduma to a broader conclusion: Broadband performance is not simply about bandwidth or headline speed.
“We’re talking about the actual user experience,” Luke Barlow says. “Today that could mean a gaming session, a Teams call, video streaming, or a live sporting event. The objective is to understand whether each person in the home is getting the experience their application requires,” he says.
Netduma says its technology is now in its fifth generation. Luke Barlow names BT in the UK and Telstra in Australia among its operator customers and says the company has also worked with NETGEAR. He also says additional Tier-1 operator contracts have been signed but have not yet been publicly announced.
The cloud knows something is wrong – only the gateway can know why
The central difference between Netduma’s approach and conventional cloud-based analytics is where and how frequently the network is observed. Luke Barlow describes a typical cloud architecture as periodically querying an agent on the home gateway for diagnostic information. Netduma instead places its intelligence directly inside the gateway, allowing it to continuously inspect network behaviour.
The company’s presentation summarises the distinction neatly: “The cloud can tell you something is wrong. The gateway can tell you why.” Netduma continuously examines traffic flows to determine whether they are healthy. That means it can detect issues such as increased latency or packet loss while an application is actually being used rather than relying on a snapshot collected before or after the event.
But context is also important: A conventional diagnostic system might identify a device with poor Wi-Fi and flag that as a problem. But that device could be a printer sitting in the basement, where a weak connection may have virtually no effect on the customer’s experience. Netduma instead attempts to identify the devices and applications that matter at that moment. Luke Barlow says the platform can then test for more than 30 different root causes of a poor experience.
Analyse locally. Decide locally. Fix locally.
Measurement is only one part of the proposition: Netduma says its software can also automatically address a number of the problems it detects. Some relate to Wi-Fi, while others can originate on the WAN side, including access-network queuing and routing issues. The architecture can therefore be thought of as a local control loop: Analyse locally → Decide locally → Fix locally → Measure the result
The principle is that when a problem occurs, the gateway is already in exactly the right place to observe it and potentially act upon it. There is no need to send the problem somewhere else, wait for analysis, and return a decision to the home. For operators, this also creates the possibility of moving from reactive troubleshooting towards proactive experience management. Instead of learning about a poor broadband experience when a subscriber calls customer support, the system may already have detected and addressed it.
Don’t send everything to the cloud – send what matters
None of this means Netduma believes the cloud is unnecessary. Instead, Barlow describes an architecture in which the gateway and cloud perform different jobs: The gateway performs the continuous, compute-intensive work of analysing, understanding, and optimising the connection. Relevant outcomes and insights can then be passed upstream.
This distinction matters because continuously sending enormous quantities of packet-level information to the cloud would be impractical. Netduma can analyse traffic locally and transmit only the information the operator actually needs. Netduma also points out that today’s broadband gateways are effectively small computers. Netduma’s software has been designed to use spare processing capacity already available in the device, rather than requiring the operator to perform all of the processing remotely. The company’s presentation describes the resulting proposition as: “Better economics. Less data. Intelligence at scale.”
Privacy by architecture
Local processing creates another potentially important benefit: Much of the underlying data does not need to leave the subscriber’s home. Luke Barlow stresses that although Netduma’s intelligence can examine packet flows locally, this does not mean all of that information is being transmitted to Netduma or an operator. Instead, analysis takes place inside the gateway and selected information can be shared upstream when required. Netduma calls this “privacy by design” and says the architecture can provide operators with QoE measurements, root-cause information, and optimisation while keeping the underlying processing in the home.
More than 10 million homes
The concept is not confined to the laboratory. Barlow says Netduma technology is now deployed in more than 10 million homes. He says the company is seeing strong NPS and customer-retention results, although detailed supporting figures were not disclosed during the interview. Barlow also cites recent field trials in which customers reportedly wanted to retain the service after the trial ended. For operators, Barlow sees several potential benefits: Lower cloud-processing costs, better visibility into the actual customer experience, more effective customer support, proactive problem resolution, higher customer satisfaction, and potentially reduced churn.
AI needs good data
The emergence of generative AI creates another dimension to Netduma’s proposition. Luke Barlow says Netduma does not intend to develop its own large language model. Instead, the company sees its role as providing high-quality connectivity data that can make existing AI systems considerably more useful. An operator’s AI agent, for example, could potentially understand that a subscriber experienced hours of successful work calls during the week but encountered several poor streaming sessions – and use Netduma’s data to explain what happened.
“We’re putting large amounts of this excellent data into them,” Luke Barlow says of the relationship with AI models. He says Netduma is already working with a major operator using the company’s data in conjunction with technology from an AI provider, although neither company was named.
This could become particularly relevant as operators introduce AI-based customer-service agents. An AI system may be able to communicate intelligently with a subscriber, but its usefulness in solving connectivity problems ultimately depends on the quality and granularity of the network information available to it.
An additive layer for the connected home
Netduma also does not position its technology as necessarily replacing the other software platforms already deployed by broadband operators. Luke Barlow says the company’s local optimisation capabilities can operate alongside other connected-home platforms, with each vendor contributing different functionality.
“The whole idea here is that we’ve essentially carved out a whole new bunch of outcomes because of our unique solution,” he says. Netduma’s intention is not to become the company that provides every function inside the connected home.
That potentially makes the architecture particularly interesting for operators that already have established cloud-based Wi-Fi management, customer-care, analytics, or connected-home platforms. The argument is not really gateway versus cloud – instead it is about deciding which intelligence belongs where. For Netduma, continuous measurement, root-cause analysis, and real-time optimisation belong as close as possible to the place where subscribers actually experience connectivity – inside the home.
And that brings Barlow back to the deceptively simple question at the centre of Netduma’s proposition: If the experience happens in the home, why shouldn’t the intelligence be there too?
/Claus.