The business of failure

10 April 2006

How do you keep investors onside when your product development slips or you need to shift strategy? And what metrics do VCs pay attention to? Neil Robertson, CEO of The Neverfail Group, explains

Neil Robertson took over as CEO of The Neverfail Group in September 2002, after more than two decades' experience in the software industry at organisations such as British Olivetti, Misys and Great Plains (now part of Microsoft). He led a major restructuring of the business, shifting direction, reshaping the board and building a new channel strategy in the UK and abroad. Now focusing on disaster recovery and high availability software and services for the Microsoft environment, the company recently secured an additional $10m funding from Fidelity Ventures.

You carried out an extensive reorganisation after joining Neverfail to refocus on the Microsoft environment. What was your thinking?

The company is 13 to 14 years' old, and started out in business continuity consulting before offering disaster recovery services in London. It did very well in the 1990s, but at the end of the decade it was becoming a tough market. In early 2002 Neverfail got a $6m investment from Advent Venture Partners with a goal of delivering a company to provide software solutions for High Availability and Disaster Recovery. But the previous management team expanded cost and overhead rather faster than it should have, and the product took longer than expected to deliver.

The company was looking for a CEO with experience of building a software company - and I was hired in September 2002. I did the normal kind of clean-up on revenue versus overhead, and in the first six months sold several parts of the business through MBOs. By June 2003 we had put to bed all the historical businesses, leaving the road clear to build a software company.

You brought in a number of new people to help with the change of direction. Getting the right mix on the board is a big challenge - how did you go about it?

I do think you need some luck. On day one I bought in the FD from Great Plains Software, who I had known for six years. He came in for three days' consulting and he's still here! We also had a big operational challenge because our product delivery process spans the globe - our development is done in Scotland, while the management offices are in Reading, we have sales support and marketing in the US, and QA is done in Romania. Coincidentally I was contacted early on by Martin Procter, who I knew from Misys and is one of the best project managers I've ever met. He joined two days later.

A lot of the incumbent management left with the MBOs, but with the existing CTO, CEO in the US, the development team, our FD and Martin I was confident I had a strong team, and we've since added to it.

But it wasn't all smooth going - your first product was late?

Yes, the product took longer to deliver than we expected - the first version was released in September 2003. We also changed product strategy. The original idea was to have an enterprise-level clustering solution that was platform-agnostic. But the problem we found was that the average sales cycle was taking nine to fifteen months. And once we did win a deal, with the market as it was in the UK, orders were often deferred to the next year. I knew this by December 2002. We wanted to get a faster sales cycle by attacking the market in a different way. So we made the product cheaper and inside the buying remit of the IT manager.

That's why we started writing Neverfail for Microsoft Exchange. The product allows a user to be reconnected within 30 to 180 seconds whenever their Exchange server fails, without any internal intervention - so we've combined high availability and disaster recovery for the small and mid-sized business (SMB) market at a very competitive price. At the end of the day, the user doesn't care how their laptop works - they just care that it does.

How did the change of strategy and product delays impact relations with the investors?

It's a case of explaining what took place and what you're doing about it. It's effectively a new market, so predicting the future is tricky - there are not many benchmarks for it. We also had a radically new channel strategy (see 'Marriage made in Heaven?'. So we walked the board through the details of the revenue projections around key metrics, and discussed the key assumptions. In many respects, we drive the company on those key metrics.

It's very unusual today for a funding round to take place and all the money to be made available upfront - most investors want to make sure the company progresses, so they might offer a third at a time. That makes for a pretty stressful time making sure that we're achieving milestones - so we have to be very clear that the goals we're setting are within reach.

One lesson I had learned from the past is that when you set expectations with VCs, some of them will allow you to get carried away and set goals that will be difficult to achieve. If you fail, there can be ratchets in terms of ownership, and if you require further funding, it can be extremely expensive.

So what measures do you use, and how good is your data?

We look at metrics such as the time from first contact to closing. In terms of marketing, we carry out all the usual forms you'd expect - branding, co-marketing with resellers, direct marketing and so forth. There's a ratio we use in terms of actual marketing spend, leads generated, leads to opportunities and opportunities to deals. We also look at the number of partners, their pipelines and conversion rates.

In the early days, with a smaller number of sales going through, the projections were less accurate. But we still captured all the relevant data from the beginning, so we can now drill into it. When we closed the investment with Fidelity, it was critical that our ability to predict the future was based on these metrics.

Neil Robertson was talking to Keith Rodgers, Webster Buchanan Research