David Cummings on Startups

Category: Tech

  • Kindle and the First 30 Pages

    There’s an interesting phenomenon that’s bound to happen with books in general and their first 30 pages. The Amazon.com Kindle, both on the specialized device as well as generic devices like the iPhone, removes the traditional friction involved in the purchasing of a book. With the Kindle, potential book buyers can read the first 20 or 30 pages of the book at no charge — why not since there’s no marginal cost for the distribution of the digital media.

    Here’s my prediction: successful book authors are going to focus even more on their first 20 or 30 pages to get readers sucked into the story so that they purchase the entire work. In the offline world, readers might browse at a local bookstore, but there is much more work involved to read 20 pages at the store to decide if the book is worth purchasing. At home, on the plane, or on the train, there’s much more down time where reading 20 pages to see if a book is interesting has less opportunity cost. With the Kindle, there will be more serious trying before buying.

    Only time will tell if this prediction comes true.

  • SaaS With No Contracts, Oh My!

    When I tell people we don’t have a contract with our software as a service (SaaS) product, they are often surprised. So many companies require a one or two year contract (much like a cell phone), that it is almost expected for mid-to-high end business software. Here’s why we don’t have a contract:

    • Tightly aligns our interests with our customers since they can leave at will
    • Requires us to get customers up and running quickly providing value within the first week
    • Helps reduce our sales cycle and establish a nice cadence with our sales team

    I recommend evaluating ways to achieve to these types of catalytic mechanisms.

  • The Importance of a Technical Co-Founder

    I was talking with a napkin stage (idea only) entrepreneur recently about his business and we got to the product portion of the conversation. He said he didn’t have a technical person on his two-person team but that they have talked to some programmers that will build the web app application. Immediately, a red flag came up in my mind. It is imperative to have a technical co-founder on the team. Why? Let’s dig into it.

    Here are several benefits of having a technical co-founder:

    • Ability to iterate faster on the product
    • The technical person can communicate directly with the potential customers or users (the company closest to the customer wins)
    • Alignment of incentives and long term interests to build a successful business

    Here are several negatives of not have a technical co-founder:

    • Shows that the entrepreneur hasn’t been able to sell the vision to a technical person to get on board and be a technical co-founder
    • Makes the project timeline and deliverables more suspect as getting a consultant’s help just isn’t the same
    • Results in more narrow-minded thinking of the technical possibilities due to not having someone on the founding team 24/7 to obsess over the future

    I recommend all teams have a technical co-founder. For Shotput Ventures, it is a requirement.

  • Frictionless Apple App Store

    For the first time in over a decade I purchased a non-business software program. In fact, I’ve purchased over half a dozen iPhone apps right over the phone. Now, I’m extremely cheap when it comes to buying digital goods (software, music, etc other than on demand movies). What aspect of programs on the iPhone prompted me to spend money? There are several key pieces involved, including:

    • Frictionless purchasing of apps
    • Effortless downloading and installing of the apps wirelessly
    • App Store ability to search and see ratings
    • Publicity on blogs and news articles on apps

    Just like Apple made purchasing digital music mainstream, they’ve done it again with buying software programs. It truly is revolutionary; try it.

  • Commercial vs Open Source

    I like to think of my company as being scrappy and resourceful. We love open source, looking to use it everywhere that makes sense. Up until 18 months ago we still had OpenOffice as the standard in the office, that was until the up-roar about incompatibility with MS Office documents became so loud that I caved and bought a license of MS Office for everyone. I wouldn’t say I’m an open source zealot, but I am a big proponent.

    My strategy has always been to start with open source and they look to commercial software once the open source doesn’t work out. As an example, we’ve used open source firewall software in our office for several years now, but it’s no longer meeting our needs. We evaluated bringing in a consultant and paying him several thousand dollars to make the changes we desired vs buying an off-the-shelf firewall appliance. The commercial firewall is being installed next week.

    My advice to entrepreneurs is to be scrappy with resources and to always evaluate open source solutions in addition to commercial systems (this from a guy who’s company makes and sells commercial software).

  • Employease: A SaaS Pioneer

    Yesterday afternoon I had the opportunity  to spend a few minutes with John Alberg and Michael Seckler, two of the co-founders of Employease. Employease, for those who grew up as part of the web-from-day-one generation, was the poster child for delivering enterprise applications over the web. What we now know as software-as-a-service has gone through several naming cycles, with application service provider (ASP) being one of the most well known during the dot com bubble.

    Employease was sold to ADP for $160 million several years ago and John and Mike are now working on their second venture called Euclidean Technologies — a hedge fund that uses advanced machine learning to do value investing in the public markets. John gave me an interesting account of just how difficult it was to do SaaS fourteen years. Here are some of the takeaways from the conversation:

    • Scripting languages like PHP didn’t even exist and their product was originally written in C and server-side JavaScript on the Netscape web server product
    • Load balancers weren’t available and had to be written from scratch
    • Internet access was spotty and would periodically go out for no reason, causing serious consternation from their customers
    • They made a big gamble, which paid off, in porting all their code to server-side Java
    • Being close to the client (a.k.a. customer-driven development) was their unique differentiator, and helped them win in the marketplace

    It was great fun listening to stories of the early days of the web, and I appreciate John and Mike sharing them with me.

  • Product Management Approach

    With a fast-growing technology company, the challenges of product management crop up on a regular basis. It is tough to balance desires and ideas from prospects, customers, partners, sales, support, services, engineering, analysts, and competitors. We’ve learned a few things over the past nine years related to product management.

    At my company, we’re always learning as we go. We recently implemented a change to all of our applications, where we put a big “Give us feedback” link in the header or footer of the major product screens. The feedback link points to an idea exchange specific to the product that solicits input from customers, prospects, partners, and employees. This approach has really helped gather ideas from a broader range of users, and it helps the best ideas rise to the top by way of the voting process.

    We also have a full-time client advocate for each product line who reaches out to customers to proactively solicit feedback. Our client advocates are awesome at doing quarterly check-ins, answering questions for clients before they need support, and being contact points to help get things done. Also, our client advocates represent one of the voices in our product roadmap decision making process.

    Internally, we have a 24 month unpublished product roadmap. Why don’t we publish it? Well, we’ve published it in the past, only to consistently have at least 20% of the items on it change. This causes consternation for clients who were anticipating a specific feature. A roadmap is very important, but we’ve learned that being fluid and agile with respect to continual feedback and market input is even more important. The world changes fast, and so do our plans.

    These are just some of our main approaches to product management. Product management is a tough discipline that requires synthesizing input from many stakeholders, and is more of an art than a science. Someone once said “the product is the marketing” and I couldn’t agree more. The product is the marketing and product management is one of the most important things any technology company will do.

  • Quick Thoughts on iPhone Development

    Last quarter we started work on two separate iPhone applications. We’d been debating for a while when to jump into it, and the launch of the $99 iPhone, along with the proliferation of iPhones among our employees, made it an easy decision. I’ve picked up on several little items related to iPhone development so far:

    • The learning curve is a bit steeper when compared to picking up a language/framework like Ruby on Rails
    • Threading is a concept lots of web developer don’t have experience with
    • If the iPhone app uses web services to interface with your own API, the iPhone app will likely push the limits of your API (we had to significantly expand our API functionality)
    • People are passionate about the iPhone and developing apps for it is great for morale

    I hope this helps if you or your company is considering building an iPhone app.

  • Thoughts on Amazon EC2 for Hosting

    We recently launched our free Web CallerID app using Amazon.com Web Services Elastic Compute Cloud (EC2). EC2 is a fully virtualized self-service hosting environment as compared to traditional dedicated hosting or managed hosting from companies like Rackspace, ServerBeach, or SoftLayer. We took advantage of Elastic Block Store (EBS) for persistent database storage and Elastic IP for persistent IPs. Here are some thoughts based on what we learned:

    • Setting up EC2 hosting takes more time than traditional hosting as you have to budget for creating/repackaging the Amazon Machine Images (AMI) and creating scripts to reconfigure items in the event of an instance change (e.g. remounting EBS volumes and reassigning Elastic IPs)
    • More or less being forced to make an AMI is a great exercise in packaging up your server for deployment, which actually saves time in the long run as you won’t have to rebuild a machine when a hard drive fails, like with traditional hosting, or when you are adding more of the same type server, like with traditional hosting
    • Amazon EC2 is actually cheaper than traditional hosting if you purchase Reserved Instances, but more expensive if you pay the On Demand rate
    • We haven’t had any maintence issues with the EBS volumes, as compared to our RAID hard drives on one of our other multi-server apps that requires a hard drive to be replaced once per month, incurring additional systems administration work on our part

    I highly recommend Amazon EC2 for hosting.

  • 3 Reasons I like Google Chrome

    I’ve been using the Google Chrome browser as my main browser for the past couple weeks and I’ve been very impressed. As a multi-year Firefox user, I was loathe to move on but the number of times Firefox would freeze on me on a daily basis was beginning to get annoying (probably partially related to too many add-ons installed). Here are three reasons why I made the switch:

    • Speed — Chrome runs a separate application process for each tab open resulting in a significant performance boost
    • Access Keys — Chrome has Alt and the access key enabled by default whereas other browsers require Alt + Ctrl and the access key (unless you manually change configuration settings)
    • Universal URL/Search Bar — Chrome has a single text box at the top that is used for both URLs as well as Google searches making it fast to enter URLs, search, and pull up previous sites
    I recommend Google Chrome for PC users everywhere.