Month 5: Figuring Out How To Sell a B2B SaaS Product

What worked and what didn't

Hey 👋

Hey there, I’m Ben, and I build things in public. With that, I hope you enjoy this post about month 5 of building startups until something takes off!

Selling RMFlags: What Worked and What Didn’t

Sales meeting in an open-plan office

A definitely real sales meeting, or something.

This month, instead of building something new given other things going on in my life like my big international move, I decided to try to do a push to sell RM Flags despite having never led a sales call in my life. Doing this kind of push required some upfront thinking about several factors:

  • Who is my ideal customer? In the case of RM Flags, I thought critically and realized a few truths about the type of customer I wanted to sell to:

    1. Despite the usual advice of “sell to startups, it’s easier,” startups actually have no use for RM Flags so that advice is irrelevant in my case. Why is this? Well, startups are too busy shipping straight to prod with no guardrails. They don’t care about feature flags or other deployment best practices and just want to “move fast and break things” as the old Facebook mantra goes.

    2. My ideal customers are in fact large, software-focused enterprise companies that are already using feature flags. Think companies like Microsoft, IBM, and so on.

    3. Given I’m selling to enterprises, I realized I needed to adjust my pricing (more on that below)

  • How should I best contact my ideal customer? I wasn’t sure of the best way to go about this so I just dove right in. YCombinator (through YC Startup School) recommends using tools like Apollo to find customers to sell to, so that’s what I tried first. It’s actually an awesome piece of software - you can set up multi-step email funnels to prospective customers with logic to automatically stop following up once the customer replies to the email. You can even tap a button to have their AI model generate emails for you! That’s what I did, and made edits to the email to make them sound less robotic before clicking Save. Here’s what my email funnel looked like in its final form:

    My Apollo sequence to get prospects to sign up for a demo

    Basically, I set up an email to send on day 1, then an automated followup a few days later, and a final followup a few days after that if they still haven’t responded. Much better than manually emailing customers myself!

    How did I find customers to fill up my pipeline? I decided to use Apollo’s tools to find ideal prospects. I mostly targeted CTOs, CPOs, Directors of Engineering at large enterprises that I knew were already using feature flags. These customers were already relatively easy to identify because I just looked at Optimizely’s/LaunchDarkly’s top customer list (they brag about their customers on their website as they should for social proof reasons) and then started emailing those companies.

    Unfortunately, this approach didn’t work at all. I had about a 5% reply rate on the email which isn’t bad, but 100% of the replies were along the lines of “never email me again.”

    Surprise, surprise: enterprise sales is a relationships business. Coming from a place of zero trust, it turns out it’s hard to get a random person to sign up for a demo from a cold email.

    I decided to make things easier on myself and message my previous company where I had already established trust with several of the engineering directors and the CTO. This went much better: I landed 2 demo meetings using this approach and scheduled 2 pilots for January. Hopefully this company will let me put their logo on my website assuming the pilot goes well, and then I can go back to the enterprise customers and try to sell them again using the social proof I got from the initial set of customers.

    Lesson learned: use your network first and get some quick wins, only start selling to “difficult” customers once you have social proof to share and have refined a pitch.

  • How do I decide how to price the product? Pricing is kind of black magic. Famously, 37signals said something along the lines of “the initial pricing for Basecamp was pulled out of our asses, and it just sounded right, so we went with it” or similar. So, I initially pulled $19/m out of my ass and put that on the website. That said, once I realized I wanted to sell to enterprises, I decided to use a more logical approach to figure out how to price my products: I decided to price according to the value the customer gets out of the product. To further explain what this means, I asked myself these questions:

    • Who is the manual process of cleaning up feature flags performed by today? A Software Engineer.

    • How much does a good mid-to-senior level Software Engineer cost per hour? Somewhere around $200-250/hour.

    • How often does a Software Engineer at a typical enterprise company clean up flags, and how long does it take them to perform this task? Once per month, and a minimum of 3 hours to make the code change, get a code review, and deploy to prod.

      That comes out to: $200/h * 3h per task * 1 task per month = at least $600

      Now, I could have priced RM Flags at $600/m immediately but I want to give my customers a good deal relative to what they’re already paying, plus I only have a basic set of features ready, so I decided on initial pricing of $190/m or ~$158.33/mo with yearly billing and 2 months free.

      Of course, I can and will raise this price later when I add more features to the product, but for now this seems to be a fair price for the value I’m providing.

      Note that the $600/mo cost to hire an engineer to do the same task is just a useful baseline for the maximum possible price I could ever hope to ask from the customer without feeling guilty that I’m ripping them off. I’ll refer back to this next time I decide to raise the price.

The Big Move

As I mentioned, I’m in the middle of an international move. I promised you all some details on the different services I used to make this move happen as seamlessly as possible, but given the move isn’t done yet, I’ll update next month on how this went instead of this month.

I will share that I’ll be moving to Thailand in early 2024, so definitely say hello if you’re in the country. Always down to meet makers, builders, pirates, startup founders, and other nefarious characters.

Either way, see you next month for another blog post on how the previous month went!


or to participate.