Voice service platforms: Interview with Dean Elwood, Voxygen

Value-added voice service platforms

I have long been a fan of the work of Voxygen, a London-based design and development company for new personal communications services. Below is an interview with Voxygen’s founder and CEO, Dean Elwood, whom I have known for many years. It captures his experience of helping mobile network operators to advance their voice service beyond basic GSM telephony. For example, Voxygen are behind services like Telefonica’s Mobile Voice Recording product, and O2 Connect (in partnership with O2 Labs) which evolved into the TU Go service. I believe it conveys the state of the art in this space, so I hope you find it of interest and value.

MG: How is the market for voice changing?

DE: There used to be three distinct segments in the voice market: enterprise, small-medium business (SMB), and consumer. The lines between these are dissolving. Already the SMB/consumer line almost doesn’t exist at all, due in part to the ‘bring your own device’ (BYOD) trend which enables consumer OTT products like Skype and WhatsApp to be used for work. Many consumer smartphone apps are equally useful in a business context. For example, Voxygen itself uses Skype for product and engineering chat groups.

The blurring of these lines radically changes how mobile network operators (MNOs) should approach the development of consumer products. It’s very hard to charge consumers for new features. New opportunities (and revenue) exist for MNOs that can create value in the work place, and charge for that value. Every consumer is potentially an SMB sales opportunity. You can replace revenue lost through diminishing returns on minutes and SMS if you seed consumers with ‘BYO-friendly’ SMB products. We see a lot of missed opportunities here when consumer and SMB departments within an MNO don’t meet or communicate regularly.

Also, SMBs are looking for ways to work smarter and more efficiently, and comms is under consistent review with all of the SMBs we talk to. Requests for better conferencing products and collaboration are two recurring themes that we see, and accordingly drive our product focus. While there is a lot out there in the space (such as UberConference), the demand we see for solutions implies none of these has quite hit the mark yet, leaving a significant opportunity. With the comms market being more than ever about mobility, this is a clear MNO opportunity.

In the large enterprise space we see an opportunity for enterprises to create richer voice experiences in their general customer interactions when exploring solutions for compliance products like Mobile Voice Recording. If you’re integrating voice recording, it’s only a minor step change (or incremental feature cost to the enterprise) to build an enhanced service, or to change current interaction models with your customer. This creates new opportunities that companies like ours are starting to exploit.

What does that untapped enterprise opportunity look like?

When it comes to accessing the contact centre, the legacy phone dialler is just filling the screen with a meaningless set of oversized numbers for users to punch at. That misses an opportunity to create a custom and differentiated experience.

However, the telco voice product design process is not joined into the enterprise customer relationship management environment, or other business systems. Since there are few relevant examples of successful innovation bridging the two, people aren’t thinking about how to re-invent that aspect of the experience. Enterprises are still trying to enhance their IVR (or are blind-sided by WebRTC rather than focusing on the customer), and not replace it with a new, native smartphone experience built on a scalable and secure telco voice platform.

What new features are proving most compelling to users?

Anything that makes existing capabilities simpler and quicker to use. Users are getting a lot fussier about design, and telcos are traditionally bad at this. That’s the nature of having a history of engineering-led product design, which is not empathetic. A really well-designed application will float to the top of its market, and the entry-level benchmark for design is getting very high as a result of the digital marketplace maturing. The telco engineering-led culture can’t easily supply design-led products centered on the user.

In terms of specific features, it is hard to avoid unified communications (UC) clichés: it is anything that streamlines the user’s communications to be more effective and efficient. Notably, many of the popular new features are based around asynchronous communications and messaging.

Consider Skype group chat being used at work. The larger the group, the more effective that capability is compared to synchronous conference calls. Another example is a security guard who can simply download a walkie-talkie application, rather than carry a bulky single-purpose device.

What are you doing to supply those new needs?

Given the importance of good design, our capability is design-led. We consider the user journey, and how to simplify it, in two different ways: the first time of use (when the user is discovering how to do a task); and how to make task completion quicker. The question we constantly ask ourselves is “what can we remove to help you get your job done better, more easily and intuitively?”. We examine how many taps and clicks it takes to get something done, and work to get it down to a ‘one click action’ (or none at all). Frequently this requires a ground-up re-design and “outside the box” thinking, plus the integration of newly-matured technologies.

For example, we have an SMB PBX-type product (TeamMate) which has a ‘group hunt’ feature. The user needs to configure the order it forwards calls to a team. We have a drag and drop smartphone UI list view to set the order people get called. Because it is a standard iOS or Android widget, users immediately know what to do. That alone is enough to create perceived new value.

A more sophisticated example is the product’s patented IVR greeting recording system. You ask the user to record an announcement of the IVR options – “Press 1 for sales, press 2 for product support, … etc.”. We then use voice recognition and artificial intelligence to automatically pre-populate the rest of the IVR management UI and system dial plan. There is nothing more for the user to do. This makes IVR creation as simple as recording the outgoing message – the product does the rest.

These are examples of the incremental innovations that can make life easier for users. There are then the ‘lightbulb moments’ and spectacular leaps forward. These come as a result of the right cultural atmosphere, and can’t be forced (for the same reason you can’t just ask a song writer to write you a hit single). So we have a focus on incremental innovation, rather than revolutionary invention. This is something that you can wrap a design process around and easily scale. Getting the right cultural conditions has taken a lot of hard work and trial and error experimentation.

How does your approach differ to how others have approach this market?

Initially, all we had was a platform, which was what we developed first and which was a hard sell. “Look at our amazing APIs and think of the brilliant things we can do together!” This forces the telco to drive creativity and empathetic design, as well as the decision about who can use those APIs to create products. That put in place too many barriers to success.

What we found is that we needed some white label products based on our own core concepts, and then to license that innovation. This allows us and the MNO to ship an enhanced customer experience right away. That tangibility sparks faster and deeper engagement with both the customer and the MNO.

Since users are familiar with different technical and functional experiences in different countries and places, the MNO then will want to tweak those ideas to suit specific markets. That means both sides need talented ‘ideas’ people: we come up with service innovation features or products, and the telcos tailor them, since they know their customers best. Even better, early customer trials of a new product enables the customer to tweak and tailor the product themselves. Usually customers understand their own problems better than anyone.

What makes us truly different is our experience of actually doing this. We have integrated our products into Tier-1 MNOs, so customers can use their existing service and number with our new features. This results in much learning, since there’s a lot of unsexy work (compliance, fraud and security, legal) that has to be put in place.

The other thing that makes us different is our track record as an “original experience manufacturer”. Our former office is Pink Floyd’s old recording studio for a reason! We have patent-protected many of our innovations, such as ‘Just Call Me’, which uses your own mobile number as a conference dial in number.

What are the organisational and process barriers for telcos innovating in voice?

The issue we have encountered working with large telcos is “who owns the success?”. It is a cultural challenge. If a large established vendor isn’t supplying the voice service innovation capability, nobody wants to own the risk. However, the moment it is successful, of course everyone wants to own it.

To get past this barrier, you need to be able to kick start at low cost and low risk. Then find one paying end-user customer to prove there is demand. A lead customer will usually be glad to partner and help you to design the product. Once you have built a credible product story you can then persuade your executive colleagues to invest more. The kind of starting money for this is under $100k. That means a $1m R&D budget can create 10 trial services, of which 8 can fail and you’ll have 2 big success stories. It’s a good model: Venture Capitalists have been doing this for years.

We have also had to learn how to deal with a telco as a supplier, and them with us. We are now ISO27001 certified, which was something we had to attain to “fit in” to the supplier model. Conversely, a heavy process-driven telco approach to OSS has required the telco in some ways to move towards our more agile approach.

What are the technical challenges you have encountered, and how have you resolved them?

You have to separate out the slow-moving telco infrastructure from the fast-moving agile service innovation. To do this at Voxygen we evolved our agile service development platform, Matrix, into a product. This can be deployed either internally within the telco, or via a dark fibre interconnect from our hosting centres into the MNO network. It powers both our own white-labellable product lines and also rapid product development on behalf of customers.

Given the continual update required by a “digital services” marketplace, a key requirement we found is the need for automated regression testing. Continuous integration is the only way to both be agile and still scale. To achieve this we have a test platform that checks out code at midnight every night, and runs tests (including UX-oriented metrics) overnight. In the morning, for example, a developer might receive an email saying that the battery life is down 4% due to a change they checked in yesterday. For many telcos that is still a manual process. What takes them 6 weeks in their development cycle, we do overnight.

Why didn’t you just become an MVNO for users with innovative voice service needs?

It is very hard for an MVNO to gain traction unless it has a household brand name with existing distribution channels. As I understand, even Disney couldn’t get its MVNO strategy right for a family-oriented product, despite being a fantastic brand that appeals to parental trust.

Someday this might change as the MVNO process commoditises. Maybe you will be able to go to mycarrier.com/mvno, take out your credit card, pay $1000, and the next day you’ll have 500 branded SIMs arrive by courier and an API to develop a product against. Today there’s still a minimum $400k to get up and running as an MVNO and, as a start-up cost to test whether the idea has any traction at all for the market segment you are targeting, that’s too steep. Fail early (which really means “cheaply”) isn’t possible.

And, in effect, you could say that we are an MVNO, but one which capitalises on the existing trust, scale, and capital of the MNO. What we bring is design talent, service innovation and a culture of agility. These are difficult for network operators to achieve internally, since these attributes typically come only in companies with a particular team culture. The operator’s structure and processes (and therefore culture) aren’t set up for making the kind of people we hire productive and happy.

What has surprised you on this journey?

A few years ago, it was the walled gardens – Skype, Viber, more recently WhatsApp – that were most successful in this space. These are ‘social enclaves’, where interoperability doesn’t matter. That led us to believe that there was little value to interoperability between networks. Consequently we put less emphasis and priority on projects like SBCs (Session Border Control) and the work that we were doing developing and integrating those. My view was that there would be less and less need for mediation and network boundary control.

Actually what’s happened with the increasing emphasis on mobile (and thereby “local”) is mediation and border control have become more important than ever. For example, if I’m viewing a search engine (say, Google) results page for the term “pizza”, then I get delivered local search results. From the search results page I can often click on a phone number and place a call to that pizzeria. I make a booking and go and spend money and eat there. How does Google get paid? How does Google prove to the pizzeria that it delivered that lead or opportunity? Here the session has transitioned from web to phone and there’s no tracking mechanism in between. Session Border Control in this context actually becomes part of the transaction flow – the most important element.

Companies like IOVOX are doing some really interesting things in this space, and we’re also working on a few things. I think it’ll be very interesting to see how the SBC companies, who also have a massive opportunity here, start to react to these use cases.

Where do you see the voice market heading in the next 2-5 years?

For consumers and SMB, we’re moving away from a mass market and towards a market with lots of niche verticals. “Bring Your Own Device” becomes “Bring Your Own Application”, and all telcos will need some kind of “voice apps store” or self-serve API to enable this.

Businesses increasingly won’t have a centrally purchased set of communications tools managed through their IT department. Users will instead choose the tool that fits the job. If you don’t sell it, they will go elsewhere.

Large enterprises will get more involved in the design of the customer communications experience. For a bricks and mortar retailer it is perfectly ordinary to regularly refresh the feel of its in-store experience. The same will begin to happen to its virtual communications customer experience.

WebRTC will accelerate this design-led voice service disruption process. There’s a simplicity to WebRTC that will make it hard to stop. Nobody loves today’s large enterprise collaboration tools, because they are a horrible experience and expensively force users to think and learn new interfaces. They are going to have to catch up with that consumer experience to stay competitive.

All these capabilities require a voice service innovation platform, which is distinct from the core voice delivery platform, such as IMS. This is similar to how there was a rush to be the dominant Web and e-commerce platform 15 years ago, in order to get more value from existing enterprise IT systems. Whoever provides this common ‘hypervoice’ innovation platform is in a really winning position in the voice technology market.

To learn more about Voxygen, go to www.voxygen.co.uk. You can contact Dean Elwood at dean.elwood@voxygen.co.uk.


For the latest fresh thinking on telecommunications, please sign up for the free Geddes newsletter.