Tech Disruptions

Posted by: Peter Carrescia | 15-09-2010

A Fast No is better than a Slow Maybe

I was reminded this week of an important lesson I’ve learned through my years in the technology industry as well as my time in venture capital.

Whether it was when I was selling software at Microsoft, managing a portfolio company waiting for a “big sale” to come in, or selling an investee company in my VC life, there was nothing more frustrating than waiting longer than reasonable for an answer from the prospective buyer. I realize too that as a venture capitalist I have been on the side of “buyer” and am very cognizant and empathetic to what entrepreneurs and management may be going through after pitching.

So while a Fast Yes is the preferred answer to anyone selling something, a Fast No is better than a Slow Maybe, as a Slow Maybe rarely turns into a Yes later. A Fast No lets you focus your energy on the next prospect, or perhaps if necessary change strategy.

In M&A, most large technology companies have professional corp dev organizations that will give Fast No’s. Good VCs will also give Fast No’s. Both groups are way too busy and see way too much stuff to have it sitting around without closure for very long.

One final comment: I have found that the best No’s are the ones that come with constructive feedback and reasoning. This doesn’t always happen, and it’s easy to understand why. Unless it is very clear that the pitch was misunderstood, an explanation of the reasoning behind the rejection is not an invitation to a debate. I can think of no situation, not one, where I changed my mind after someone debated my logic. Best to take the feedback, incorporate it if it makes sense, and move on. You can always come back later if you’ve made progress.

2 Comments

Posted in VC

Posted by: Peter Carrescia | 14-09-2010

Customers vs. Prospects

In over 10 years as a VC I’ve seen hundreds of companies pitch in the hopes of raising capital, and I’ve seen thousands of business plans.

A significant chunk of those companies were pre or early revenue, and I can understand the desire to appear further along than one may really be (particularly to potential investors), but so often I see the term ‘customer’ used as a synonym for ‘prospect’. To be clear, a customer has already purchased something from the company, a prospect will hopefully buy something, but has not yet done so. Obviously, a pre-revenue company does not have customers, they only have prospects.

Sometimes though, beside fundraising truth-stretching, I think some companies don’t see the difference. I bring this up because there is an important distinction, particularly with early-stage companies that are still defining their product/services. I have seen countless of “If it was on Linux, on Windows, avail on iPad, avail on Android, uses Java, uses VMware, did this, did that, then I’d purchase it”. Customers tend to only ask for things they really need to make your product work better. Prospects don’t often have the same insight as customers nor the same motivations. And unfortunately, often the feedback is filtered through a sales person who may not be able to identify the motivations behind feature requests.

In early-stage companies, the feedback received from customers and prospects is often funneled through the CEO (vs product management in later-stage companies). Great companies can pick out the key requests of customers and prospects, identify the motivations behind the request, prioritize it, and then ensure that resources are allocated appropriately to meet those requests. Not-so-great companies will jump from request to request with equal urgency, wasting time and resources as their competitors better address their market and in a more timely fashion.

Outsourcing Development

Over my years as a VC, I’ve been involved with a number of software companies that have, for a variety of reasons, chosen to outsource some subset of development. The decision to outsource always made sense at the time, and almost always had to do with ‘time to market’ issues. Looking back now though, in almost all cases the decision was later regretted. The few times it did work effectively, there were some common attributes. First, the outsourced functionality was discrete and not at all contingent on the core development process. Second (and somewhat related), the outsourced code interfaced with the core code through a defined, well documented and very stable set of APIs. And third, the outsourced code did not include any “user interface” or “user experience” code. Because of my experiences, I think I’ve got a pretty good idea of where and when outsourcing can be effective.

I was reminded of my aversion to outsourcing yesterday as I was installing my new Cisco/Linksys E3000 router. This router is part of a new line of routers that Cisco announced earlier this year. I’ve been using Linksys routers for years and am quite familiar with the web-based management console, and it only took a few minutes to configure the router to my specifications.

One of the selling features of this new line of routers is a new user-friendly management console called “Cisco Connect”. Cisco Connect is an application that is installed on a PC and allows you to manage the router through a rich application that hides a lot of the complexity of the web-based console. While I don’t know if the development of this application was outsourced by Cisco, it certainly has all the hallmarks of what goes wrong when you outsource the wrong kinds of development away from your core development team.

I didn’t initially install Cisco Connect as it was positioned as a product for non-technical people and a subset of the functionality that was available in the web-based traditional console I’m familiar with. Well, it turns out there is functionality that is only available through the Cisco Connect application (guest accounts, richer access control functionality). So last night I installed Cisco Connect so I could use this functionality. On attempting to launch the application, it terminated, stating that changes had been made through the web interface and that Cisco Connect could not be used.

So I then reset my router and retried Cisco Connect. This time Cisco Connect did start up, but when I went to make a few very basic changes to the router settings, changes that Cisco Connect defines as ‘Advanced’, the ability to continue to use Cisco Connect is disabled.

Use the traditional web interface and don’t get access to key functionality. Use Cisco Connect, and don’t get access to key functionality. Dont use Cisco Connect, and you have a guest wireless network broadcasting its presence that can’t be turned off. and if you choose to use Cisco Connect, it forces you to make poor security choices such as SSID broadcasting and common passwords across the wireless security and the management console. And crazily enough, there is no password needed to launch Cisco Connect, and while in it, it will offer up to you the management password (and by extension the wireless password as it forces them to both be the same). And in a final affront to security best practices, it stores the security password unencrypted in a plain text file on the computer. That you have to choose between a subset of the router functionality and full security, or some other subset of functionality and poor security is unacceptable in any case, but from Cisco?

Which brings me back to outsourcing development. I don’t know if Cisco Connect was an outsourced project (perhaps it was developed by a recent acquisition?), but it certainly has all the hallmarks of outsourcing gone wrong.

Biting off more than you can chew…

I recently wrote a blog piece entitled “This is not a Blackberry-only Issue” on RIM’s security problems with a number of countries. In it, I commented that:

…the capabilities that RIM provides have been available in off-the-shelf software for over 20 years now. There is nothing that RIM provides that cannot be put together with countless other email and secure instant messaging solutions… [Subversive groups] will now simply use web-based mail like Gmail from a mobile device with SSL turned on.

Since the time I wrote that, it has become clear that India has been far more encompassing in their demands. In addition to BIS, PIN and BBM messaging, they are also demanding access to all emails, including corporate email between employees that use the BES server (an outrageous requirement for a business-friendly country and, assuming companies use secure VPNs, one that they don’t have today whether a company uses Blackberries or not).

Last week, RIM smartly brought the rest of the tech industry into a fight that they were happy to stay out of when they suggested in a statement an industry-wide forum to address the issues that strong encryption introduce to countries that are lawfully monitoring communications. In the statement they said:

The use of strong encryption in wireless technology is not unique to the BlackBerry platform. It is unquestionably an industry wide matter… Similarly strong encryption is currently used pervasively in traditional VPNs (Virtual Private Networks) on both wired and wireless networks in order to protect corporate, government and law enforcement communications… Singling out and banning the BlackBerry solution would be ineffective and counter-productive.

The statement went on to conclude:

…as India continues its path of strong economic growth and increasing international business, the need to protect the integrity and security of sensitive corporate information through strong encryption-based information and communications services will grow. Banning such strong encryption-based information and communications services would severely limit the effectiveness and productivity of India’s corporations.

Apparently, someone in the Indian government actually understands the problem, but instead of backing off as they should, they’ve decided to step up their demands, as today apparently they are turning their sights on other companies (see Google, Skype under fire in India security crackdown).

Let me predict now that India will back down – in this case they’ve bitten off more than they can chew. It is one thing to pick a fight with RIM. It is quite another to pick one with Google and other US companies such as Microsoft and Cisco (providers of strong encryption software, services and communications equipment). India’s economic strength, and in particular it’s status as a key branch office and outsourcing provider to US companies, will be in jeopardy if US companies cannot trust that their communications to/from their Indian offices are secure. And US politicians will use any excuse to restrict the ability for US companies to move jobs to foreign jurisdictions – passing laws that restrict the ability for US consumer data to be transmitted to countries that don’t ensure protection of that data from foreign government prying eyes is an easy one to sell to the public (heck the public doesn’t trust their own government, there’s no way they’re trusting someone else’s!).

All RIM needs to do now is find a way to let India back down gracefully, allowing them to declare victory even in the face of defeat.

Paying for Intellectual Property

I saw this humorous post on Gizmodo yesterday entitled “This is How I Feel About Buying Apps” (see original post by comic artist Matthew Inman here).

People have no problem spending $5 for a coffee at Starbucks, $200 for concert tickets, or $60/month for their smartphone plan. But ask them to spend $0.99 for a song or an application and they will either steal it in the case of a song, or not purchase it in the case of an iPhone application (there’d be a brisk business in free app-sharing over P2P too if there was an easy way to do it).

I am amazed that kids listen to the same 10 songs for months at a time, and have iPhones, Blackberries and Android phones that cost hundreds of dollars, and with plans that cost hundreds of dollars annually, yet steal what would only cost perhaps $100 per year. I remember as a teenager going to the record store weekly and buying albums (and then CDs) for $8 and then up to $18 when CDs were introduced. I did that almost every week as a teenager! And there were often only a couple of good songs on any one album.

Think of the price erosion in the music business: 25 years ago we happily and regularly spent $15 for a CD with just a few songs we really wanted. Today, we only have to spend $3-4 to get the same songs without even having to get in a car and go to a store, yet we go out of our way to find a way to get them for free.

Why is that? And I guess, how do business models change to adapt to this reality? As an investor, how do you invest and successfully build a businesses that sells digital (aka. easily stolen) goods?

I guess with $200 concert tickets, I know part of the proposed answer to that.

1 Comment

Posted in General, Software, Strategy

Tech Disruptions

A Fast No is better than a Slow Maybe

Customers vs. Prospects

Outsourcing Development

Biting off more than you can chew…

Paying for Intellectual Property

Categories

Recent Posts

Email Subscription

Archives

Follow “Tech Disruptions”