BT & Ribbit: Progress Report

We’ve been following Ribbit and their efforts to create a platform for rapid development of voice & messaging applications for two-and-a-half years. The reasons why we were interested are articulated in this quote:

Further in the future, though, highly reconfigurable telephony is likely to lead to radically different product and business models for telcos. For example, civil engineers stamp out custom bridges off well-tested models based on span, load, and topography. Your telco consulting services arm will be building custom communications experiences, with the software equivalent of a flexible manufacturing system. Custom back-ends, process flows and user interfaces will be generated from tools and models. Each is created appropriate to the application and user context. Most devices will have a completely “soft” and re-configurable user interface.

When BT acquired Ribbit in the summer of 2008, Thomas Howe pointed out in a guest post that CEBP projects typically result in a productivity gain of around 20%, and 20% uplift in productivity across the whole economy is a lot of money by anyone’s standards. We had a look at some different CEBP/Voice 2.0 business models in this analysts’ note, and discussed the technology strategy aspects.

So, how’s the business coming on?

Many people felt BT had hugely overpaid for the company in laying out $105 million for “100 developers working on salesforce.com”, as one Oracle executive described it. As we pointed out here, though, it had within it the potential for a transformation of BT’s business – a strategy that would integrate developer APIs, open CRM systems, call centres, and broadband connectivity, across BT Retail, Wholesale, and Global Services, as Martin Geddes put it.

That was the theory; Martin, of course, left Telco 2.0 to take part in this transformation himself at BT Design. However, the politics and economics didn’t work out. Ben Verwaayen’s departure from BT seemed to rob the initiative of top level support; the recession revealed the downside of BT Global Services’ rapid expansion, as the division turned out to be sitting on an impressive pile of bad contracts. The doomed NHS National Programme for IT contracts were almost enough to add up to a major crisis in themselves, and the plummeting stock market caused BT’s pension fund to need a huge cash injection. BT shed some 35,000 jobs, Mohamed Al-Mojli and Martin left the company, and J. P. Rangaswami was transferred from BT Centre to take over Ribbit personally.

Despite all the drama at the corporate level, though, Ribbit has been quietly achieving. The launch of Ribbit Mobile, its first consumer (well, prosumer) product, got a lot of attention and favourable comparisons to Google Voice, like this CNET review, or this story on TechCrunch:

Ribbit Mobile’s iPhone app is fine as far as it goes, but I kind of just want it to take over the entire phone function of the iPhone, which is the part of my iPhone I use the least anyway.

Let’s repeat that: I kind of just want it to take over the entire phone function of the iPhone…which is the part of my iPhone I use the least anyway. One sentence sums up the potential of better voice and messaging, and the threat to the 66% of Vodafone’s revenues that consist of voice or SMS. Here’s an example: Caller ID 2.0, a feature for Ribbit Mobile that links your phone number, your address book, and LinkedIn’s API. Ribbit’s recent developer challenge shows some fascinating examples of the creativity open telephony enables – Sena Gbeckor-Kove integrated the open-source CRM system Sugar, the winner integrated telephony into an Augmented Reality application.

That’s all very well, but it’s admittedly a little shiny and tech-newsy. That said, it’s absolutely essential for a business like Ribbit that it gets attention – a developer platform without developers is fairly pointless, and Ribbit has so far signed up over 20,000 developer accounts. The last time we checked in the count stood at 600.What about the strategy? Is BT still aiming to fix Global Services with open-source software and developer APIs, as J. P. Rangaswami said when Telco 2.0 visited BT’s open-source development shop? We had a chat with Crick Waters of Ribbit to find out. Here are some of the key points.

Ribbit’s biggest customer: BT

The biggest customer for Ribbit is BT, in three different roles. BT Global Services is making heavy use of the platform to build applications for its clients, which allows Ribbit to leverage the BT sales force and customer base, while the BT GS consultants and developers get another powerful tool to work in parallel with Salesforce.com, Sugar CRM, Avaya, Microsoft OCS, Asterisk, and friends. The very simplest demonstration of this is on the main BT website for business customers in the UK – ask them for Salesforce.com and they’ll offer you the Ribbit plugin and Ribbit Mobile.

BT Global Services deals in key accounts and giant contracts – multinationals, global carrier services operations, government contracts. By contrast, BT’s small and medium-sized business products sit within BT Retail, in a division called BT Business, parallel to the traditional residential operation in Consumer.

At their 2010 investor day, BT executives made very clear that the key strategic priorities at BT were as follows: executing on FTTX deployment, relaunching BT Vision IPTV, aggressively promoting networked IT products to SMBs, and providing communications and IT solutions for their key accounts among global MNCs and government. Three of these are outside the scope of this note.

However, the third is very much in scope – BT wants to defend its market share here by introducing advanced features that rival operators don’t have, and to replace declining voice margins and volume by upselling its existing small business voice and IP customers to richer networked-IT/cloud computing services. Ribbit is crucial to this strategy; its fundamentally network-resident nature means that BT can deliver better voice and messaging services to any business that takes its service.

Providing CEBP for businesses ranging from sole traders to FTSE250 enterprises (BT’s definition of SMBs) is a significant operational challenge. There are obvious costs to delivering on-premises or even traditionally hosted Asterisk or similar products to hundreds of thousands of SMBs. If BT can provide a similarly capable platform as an on-line service, it can deliver it through its on-line provisioning workflow, as it does Internet service, wholesale Ethernet feeds, or telephone calls. Rather like Telenor Objects does for Machine-to-Machine applications, it’s a platform that horizontalises CEBP and unicomms for Person-to-Organisation applications and delivers them from the cloud.

Since we spoke to Ribbit, BT has announced the full integration of Ribbit into its SMB VoIP product, One Voice – essentially, Ribbit is now BT’s premier offering to business in its home market. BT is aiming for the “doughnut hole” we identified between the large corporates (and tech-focused smaller companies), for whom running their own Asterisk infrastructure and developing applications for it is well worthwhile, and the hyper-simple sole traders whose IT needs are met by consumer-grade products.

Another role is as a productivity tool for BT internally – there are a lot of telephone calls going on within a company the size of BT…

Staying Relevant: Injecting Innovation Into the Core Network

It’s worth remembering that Ribbit’s assets weren’t limited to the code that implements the Flash API – there’s a whole Class 5 softswitch of theirs underlying it all. Fascinatingly, BT is increasingly making use of Ribbit technology in its core network, as it begins to grapple with the question of what will replace the PSTN. According to Crick Waters, it’s a question of how operators remain relevant as providers of real-time communications in a multi-screen, multi-device environment, where “everything is a computer” and dial-tone is no longer enough. He named convergence, multiple identities, single sign on, and switching between asychronous and synchronous modes of communication as critical fields.

Ribbit co-founder Ted Griggs has been given the newly recreated title of CTO Voice in BT itself – readers with long memories may recall that BT once had a post of Head, Global Voice Technology, which was then abolished. Ribbit developers are working on BT-wide applications, and Griggs is apparently spending significant amounts of time on 21CN business – this would seem to show that there is again real commitment to better voice and messaging at BT.

Click-to-Call, Better Signalling, Voice to Text

The biggest Ribbit customers by sector are digital advertising agencies, who are heavy users of the click-to-call API, which lets them integrate real-time business processes with their ad campaigns and collect richer information about the potential customer and their interaction with the ad before the call begins, and then financial services companies. FinServ firms are especially interested because of the Sarbanes-Oxley Act compliance requirements, and also for productivity reasons – typically, in many parts of finance, much business is done on the phone in unstructured format.

Ribbit applications are used to capture telephony, convert the voice stream to text, and archive it – this provides SarBox compliance, but also a source of structured data about the business’s interactions with its customers. Not only can you pass more data about a call to the call taker, or the call originator, when it originates from a Web application, but you can also pass data about the call to an application for automated processing – but only if you’ve got the metadata and you’ve got the conversation converted to text. Ribbit expects an arms race in voice-to-text applications.

Rolling Out with More Carriers

Ribbit is planning to expand rapidly by signing up more telcos to offer the service. Evidently, the signalling/control side of Ribbit is global – it’s a Web API. The voice side, however, isn’t. The answer is for new operator partners to hook up their networks to Ribbit’s through SIP peering – Ribbit calls this a “bring your own network” model, and they are interested in integrating Telco 2.0 capabilities such as location-based services and internal CDNs, so that developers working with Ribbit can get them through the Ribbit API. One advantage of the SIP interconnection model is that it permits very quick deployment.

Crick Waters argues that Ribbit’s core skills are “engagement, metrics, and monetisation”, and that they are working with a number of operators on their API strategy. Smaller operators are the focus of their deployment push.

Some Further Thoughts

It’s notable that, as well as cross-selling Ribbit into the UK business customer base, BT is also cross-selling BT Global Services products into the Ribbit customer base – for example, the Ribbit Web site is currently advertising hosted contact centre services, a key BTGS line of business.

Overall, BT’s strategy will tend to spread the profits from better voice and messaging and their integration with Web applications across BTGS’s activities with contact centres, multinational companies, and government, and also across the BT Business element of the BT Retail division – Ribbit’s relationship with BT is that of an internalised supplier of technology.
Successful execution of this strategy will be very important for BT in the future, and for Voice 2.0 in general.

BT is thinking about the future again, after the organisation went through a wave of introspection and loss of confidence; one sign of this is that they are considering what kind of customer premises equipment might be the future gateway for operators into the home. At the 2010 investor day, they demonstrated an iPad-like tablet device – others might think a femtocell, a Google TV, or a media-server device providing content and home-automation interfaces to mobile devices around it more likely. What matters is that they are experimenting, rather than retreating into denial or getting stuck in option paralysis.
Long before that, they are clearly aiming to use Ribbit’s technology and developer community model to become the primary supplier of IT services for UK small businesses.