Voip2020

Fun-draising?

I am now deep in the process of raising Toktumi's next traunch of capital. We raised $1.5M last year from various angels and strategic investors, and have managed to launch the service while keeping our burn down and our operations very lean.  But after our successful launch at DEMO, we need capital to build our management team, accelerate development of new features and services, and upgrade our infrastructure.  I've been very pleased with our traction - more than 10% of our free customers have converted into paying customers, which is higher than I expected given that we are still in closed beta.  We are also selling toktumi phone adapters from our webstore despite no promotional efforts.  I think we've definately tapped into an unmet need.  I've also been encouraged to see our existing customers invite new ones - so far there's been, on average, one invite sent per acquired customer, which seems to indicate the potential for some viral growth.  When we marry that with deals in the works with channel partners that already reach small business, that should help us reduce our customer acqusition costs.  I'm convinced that our approach is the only profitable way to acquire customers in SOHO 1-9 employee businesses, which have traditionally been underserved because its been too expensive to reach them selling complicated equipment via call centers and direct sales teams.  By offering an easy do-it-yourself system for which referrals can quicky lead to trials, we are the first to really unlock the power of web distribution in the hosted PBX market.  Now, let's hope the VCs can see it the same way!

The Sting of Fraud

It only took them a week. I thought it might take longer, but its amazing the speed with which fraudsters with stolen credit card numbers found Toktumi.  We were able to shut down the fraudulent accounts quickly and permanently block the PCs from creating new accounts, but it hurt to see it happen so fast.  We refunded the charges to the stolen cards and informed the cardholders who contacted us of the situation and warned them they needed to contact their banks immediately.    We've now tightened security with a variety of methods that I shall not reveal. 

Telecom and fraud have a long history and despite the best efforts of Cybersource, our credit card authorizer, bad cards still got through.  Its up to providers like us to develop our own proprietary ways to protect ourselves and customers from fraud.  Over time we'll get better and better at stopping them.  Its part of the IP that is created in the course of doing business online.  Although it stung to be hit with fraud so early in our history, I know that now, while we are still in our beta, is the best time to do our learning and refine our approaches to reduce its impact on our business.

What is Viral?

I had a very interesting conversation with Mike Cassidy last Friday.  For those of you who don’t know him, Mike founded three successful startups, the most recent being Xfire.  Xfire is an application for gamers that grew 2% a week to 2M users before it was acquired by MTV Networks in 2006.  It now has over 9M users.

So what about Xfire made it viral?  Mike explained that an application can only be viral if it requires others to use it to make it useful to you.  So hotmail was clearly viral, as was Xfire, because it allows you to monitor and communicate with your gamer friends.    You need your friends to use it to make it useful to you – so you spread the word and get others to download it.  That’s what makes it grow.

But that’s not enough. You still need some math to work to achieve exponential growth, which is when a viral application takes off.  Here’s the math, as explained by Mike, who attributes this to Ed Baker, a 2nd year Stanford b-school student:

X = % of people who accept invitations and activate an account

Y = % of that group that invite at least one other person

Z = Average number of invitations sent per user (of those that invite at least one person) in the period of measurement

Viral Exponent = X * Y * Z

If your Viral Exponent exceeds 1, then you’ve got a viral application.  And it doesn’t need to exceed by much.  Xfire’s exponent was 1.02 based on a one week period.  That meant it grew at 2% a week.  Knowing that, Mike could predict exactly how many users he would have – and his predictions were right every month over three years!

Toktumi is alive and kicking

Well, the company is launched.  Its been going surprisingly smoothly - I'm waiting for the shoe to drop - I know there will be drama at some point - but so far it was a great week. We've gotten good coverage (see www.toktumi.com for some of the articles and to see our demo) and made at least three "top picks" lists from folks covering the conference.  We've had a fairly steady stream of people signing up for the service (although it slowed today going into the weekend).  Our best signup day was Thursday.  We are seeing existing customers invite new ones, which is encouraging.  We made our first sale on our first day, which I hope was a good omen.  We've also had customers convert to the full paid service already, despite our still being in beta.  So all in all, a good first week.