Principles behind a Freemium Pricing Model

It seems that there are only two ways you can go about defining a pricing model for your web app these days. Either you blatantly copy your competitors (or 37Signals if you don’t have any direct competition), or you juggle through different permutations until you get to something that “feels right.”

While thinking through a pricing model for BrainTrust, after numerous failed attempts (including the one shown on the site right now), I decided to write out some principles first.

  1. There should be a free plan to give the casual user a taste of what we are about. Anything given for free should only give a taste, at a clearly limited scale. Not all features should be given out in the free version.

  2. There should be a “recommended” plan. This plan should cost what we need the majority of our user base to be paying. It should also have the appropriate features to give the BEST OVERALL EXPERIENCE our product has to offer.

  3. There should be a “cheap” plan for the skeptics, that includes all of the essential features but at a limited scale. Remember, we’re only giving a few limited features in the free plan, skeptics can upgrade to the chapo plan to get a feel for all the features. However, the cheapo plan still doesn’t provide the “best experience” — thereby giving the customer an incentive to upgrade to the recommended plan once their concerns have been alleviated.

  4. Finally, there should be a “hail mary” plan. This is the fully equipped, no expenses spared version of your product at a triple digit price, to test out the upper limits of our market and our overall pricing assumptions.

  5. Our customers should be able to take any of the paid experiences for a trial run. The trial run should be long enough for the custimer to run through all the important use cases the product supports.

  6. There can be “filler plans” that give our customers a healthy gradation to pricier plans — so that our customers can slow upgrade to larger scales without having to jump straight to the “hail mary” plan.

I hope this helps. I’ll be sharing how our new pricing strategy works out for BrainTrust once we go live.

In the mean time, comments, thoughts. debates are all welcome.

  • Pingback: startupbug.com()

  • Pingback: iPad Links: Monday, March 1, 2010 « Mike Cane's iPad Test()

  • Pingback: 9 principles behind an effective landing page | Tawheed Kader (TK)()

  • I think these are good rules (though we don’t follow any but #1)… Seems like it might make for some artificial complexity. Adding filler and hail mary plans seems to add a lot of noise. I’d A/B test that.

    Also, for your FREE plan, I’d suggest giving them ALL PAID FEATURES FOR 14-DAYS. That’s what we do. Of course, you encourage them to sign up for a paid plan, but if they don’t, say “Hey, we’re giving you paid features for X days anyways”. If your pay wall is in the right place, a meaningful % of these people will convert to paid instead of lose the features they’ve grown to love.

    Stepping back, I think your biggest problem is positioning– starting with your home page. Your position is entirely feature-centric rather than benefit-centric. As a small business owner, WHAT ARE YOU GOING TO DO FOR ME?

  • While what you describe fit many of the pricing plans seen across the web, I believe there needs to be more rigor and analytics behind designing versions and pricing. There are costs to customers in choosing the versions. The startup must also understand the value of different versions to different customers and price them such that those who can and have high willingness to pay will choose the higher priced version and not tempted by the lower priced version.

    Pricing and versioning should be strategic decisions driven by customer needs, segments being targeted and alternatives available to customers.

    If rigorous customer segmentation analysis is out of reach for a startup the one principle I recommend is Hal Varian’s Goldilocks Pricing.

    -rags http://twitter.com/pricingright

  • Pingback: Updated Pricing Plans | Braintrust.io()