Author: David Cummings

  • Rapid Improvement with SaaS = Reference Customers

    Another SaaS benefit that isn’t readily apparent before going live is the relationship between quick product enhancements and reference customers. Because you can update and enhance the product so fast, customers that make small requests can see the changes within a matter of days or weeks (assuming you accept the tweak and it fits in your opinionated vision!). This goes a long way towards having them become a reference account that you can use for future prospects. With installed software, it is difficult and time consuming to keep making little patches so customers often have to wait several months for the release cycle to happen before they can actually appreciate that your team has been working hard on their request.

    Near instant gratification with product change requests is something many clients have never experienced. Please make sure and temper expectations that it won’t always happen but that you’ll always be there to hear them out and be understanding.

  • New Product Pricing

    Pricing for a new product is really just a shot in the dark. What features go into what editions? What are the up-sell hooks? In my experience it is much easier to lower prices than to raise them. I’d recommend starting out with a price that is twice what you feel is right. Why? It is very easy to give “preferred customer discounts” and come down on the pricing, making the new client feel special. Throw in membership to your “product advisory council” and you’re golden.

    Another huge consideration: credit card price range or CFO price range. This is a big deal. Having a buyer be able to put the product or monthly fee on their credit card is a much easier sale than one that requires a purchase order and needs sign-off from the CFO. Wait, aren’t there plenty of deals between the $1,000 (credit card) and $50,000 (CFO sign-off) range? Yes, those are available but you don’t want to play in that space. The amount of time to complete a $20,000 deal is the same effort as a $50,000 deal. If you have to do a full-service, long sales cycle process, make sure your package is at least $50,000.

    My final thoughts: don’t spend much time on pricing, go to market quickly, and listen to your prospects.

  • Atlanta CEO Council Breakfast with the CEO of Red Hat

    The Atlanta CEO Council put on a great event yesterday morning. As part of their guest speaker series, they brought in Jim Whitehurst, the CEO of Red Hat, to share his vision as well as lessons learned being the COO of Delta Airlines. After hanging out with Lance Weatherby of the ATDC and Andy Monin of VendorMate, the presentation began. Here were Jim’s key recommendations for CEOs:

    • Focus, focus, focus – he mentioned three things they focus on for a year and work hard to say no to suggestions that don’t fall in those categories
    • Build repoire in person with all lines of company employees – he talked about meeting with 12,000 flight attendents over a six week period (350 at a time) as a way to help with customer service at Delta

    It was a great event.

  • Use Your GPA to Know When to Expand

    As a software entrepreneur it can be difficult to know when to expand staff, marketing, etc. One simple method I developed over the years is called GPA (Growth Plan Assets). The method is really simple in that you add together cash in the bank and current accounts receivable and then divide by last month’s “normal” costs. This, in a rough fashion, gives you the number of months you can operate without any new sales.

    You know you’re ready to expand when your GPA is greater than a standard college GPA (scale 0 – 4). So, like a college GPA, most of the time you’re in the two or three month range. When you go above that, you have a sufficient GPA to expand. When you are below a two, you’ll have some tough decisions to make. What’s your GPA?

    Note: The higher the percentage of revenue that is recurring, the lower the desired GPA.

  • Thoughts on Domain Names

    What’s in a domain name? We’ve spent a fair amount of time lately debating the name of our new product as well as the importance of a domain name to go along with it. Here are the domain name thoughts so far:

    • Keep it short
    • Include the product name in the domain name
    • Use real words (e.g. prospectinsight.com) unless there is one that is short and memorable (e.g. pardot.com)
    • Get a .com and don’t settle for another extension
    • Look at the domain name auctions and after market

    It is extremely hard to find a good domain name these days. Good luck!

  • Complex Sale or Order Taking Sales Organization

    Do you have a direct sales team that works hard selling or do you have an order taking call center and website? That’s an important distinction when building a business. With a direct sales team, you have the following characteristics:

    • Bigger deal sizes
    • Longer sales cycles
    • Complex sales cycle and pricing model
    • More expensive personnel

    When sales come from an order taking call center and website, you see these characteristics:

    • Larger marketing budget as a percentage of revenue
    • Smaller deals
    • Shorter sales cycles
    • Free trials as well as “free” editions

    It is important to consider both of these options when thinking through the possibilities of a new business application.

  • Products Roadmaps

    Does your company make product roadmaps available? I’ve always tried to avoid providing product roadmaps for prospects and customers for several reasons:

    • Agile products need to adapt on a monthly or quarterly basis
    • Similarily, the market changes too fast to know what will be important in 12 months
    • If they see something they want and it doesn’t get done they won’t be as satisfied (better to under promise and over deliver)

    Instead of providing a roadmap, find out what is important to your prospect or customer and make sure it meets their needs now. Sure, it won’t be 100% of what they want but it is better to provide real value now than to hope it will be ready in the future. Hope is not a strategy.

  • Always People, Process, and Technology

    Whenever an issue is brought up to me, I always like to think of it in terms of people, process, and technology. In the two markets that I’m intimately involved with, web content management and web marketing automation, we try to stress that it isn’t just the software or technology that will result in a successful client.

    • People – are the right people involved?
    • Process – how are things currently being done?
    • Technology – what tools and applications are involved?

    This simple framework goes a long ways.

  • Productizing Services Work

    One of the best moves we’ve made for our consulting and services teams is to “productive” the work. What does that mean? That means that instead of quoting project work as a “bucket of hours” (e.g. 100 hours at $200/hour for $20,000) you have different “products” available like a Professional Quick Start Package that gets a client up and running, or a Newsroom Package that delivers a certain number of templates, pages, workflows, etc.

    Productizing the services work aligns interests from both the client and vendor to focus on the benefit and outcome of the work — not how long it takes. This took a long time to realize, and we’re not completely away from quoting hours of work for certain projects, but the majority of consulting work around a product should be productized.

  • Example B2B App SaaS Infrastructure

    After our hosting provider had internal network issues last week, setting up redundant datacenters became a top priority. After much debate, here’s what we decided on for our new SaaS infrastructure:

    • Enterprise-class round robin DNS with automatic fail-over
    • Two datacenters with identical infrastructure:
      • One Linux load balancer (1 dual core CPU, 2 GB of RAM)
      • Two Linux app servers (1 quad core CPU, 4 GB of RAM)
      • High-end MySQL database server (2 quad core CPUs, 12 GB of RAM, RAID 5)

    The database servers are configured in a master-master replication setup such that each datacenter is fully autonomous.

    Now, time to get to work and set things up.