
The only way to get good at something is to suck at it first.
Here are all of the mistakes weโve made over the past two years building Confluence.
Hopefully at least one thing in here will help you avoid mistakes in the future.
Stopped creating pipeline and settled for inbound leads.

Our main marketing channel is word of mouth (at the time of writing, 97% of members come through a referral). We track this through ourย application, and we showcase it on ourย website. The issue with settling with this as the primary customer acquisition channel is that you leave a lot of demand on the table by not doing any outreach.
LinkedIn gets trashed by a lot of people in tech, but Sales Navigator is a cheat code for sales teams. Gets Sales Navigator, build a free lead magnet, message like a normal person and not a salesman, and youโll be shocked at what happens to your pipeline. If you need help mastering your outbound,ย check out this.
Gave away too much for free
Our community has always been invite-only, but we only switched over to a paid model in October 2021. We were doing everything for free for the first 15 months of the business, and this set an expectation that we were running a non-profit instead of a business. If you are running a freemium product or a free project that you plan to monetize later, Iโd recommend readingย thisย to help you think through the pros and cons of the approach.
Building for non-paying customers
Related to the point above, a big mistake we made was adding additional features for the free members before planning how to recoup costs and monetize later. If a customer is not paying you, they are not a customer. Avoid situations where you create an expectation that you are willing to work for free. Figure out how to monetize, then build for those that are directly supporting your business. This is the only way to align incentives.
Prioritized low-leverage activities
One of the biggest mindset shifts Iโve taken over the past year is to focus on activities that bring in revenue. Growing the member base, increasing reach through more content, expanding partnerships โ these all increase revenue and directly impact the business positively. Everything else should be outsourced.
Writing copy, creating visuals, prepping questions for podcasts, scheduling guests, scheduling newsletters โ all of this can be taken off my plate, but I havenโt done that. Iโm in the process of finding an EA that I like working with, and I expect to offload more of this work to her. If youโre looking for a good EA, I recommend finding one onย Upwork.
Did not prioritize audience building
This goes back to the first mistake we made. The world today runs on attention. Especially when youโre starting out, building your reach and authority should be a top priority.
If you find something that works, you have two options. You can either accept slow growth by continuing to do what is working, or you can pour fuel on the fire by doubling down on what is working. I still hate the thought of being self promotional, but if you donโt promote yourself, nobody will.
Your options:
โ @jackbutcher (@jackbutcher) April 10, 2020
1. Promote yourself.
2. Work for someone who promotes themselves.
Tried to sell to existing audience rather than expand to new audiences
I originally thought that the path to wealth was building a massive audience and selling directly to that audience. That works for some, but Iโve come to realize thatโs not the game weโre playing.
Weโve soldย digital productsย andย tangential subscriptionsย to our audience, but this hasnโt been a killer success for two reasons. The first reason is that nobody likes being sold to, and the more sales-y we got, the more pushback we received. The second reason it doesnโt is because the price of the products we were selling did not justify the work being done to generate the revenue.
Knowing what I know now, Iโd rather spend my time creating crazy amounts of value through better products and better content, and have that help us grow into new markets. The talk below shaped my thinking of how to think through audience, monetization, and playing long-term games.
Did not solve for bottlenecks
This is a regret because it is still not solved for.ย Our subscription model is intentionally not scalableย because every member accepted into Confluence is vetted and approved. Compare this with other software companies that want make it as easy as possible for any website visitor to become a paying customer.
This has limited what we can do and who we can sell certain products to, but I ultimately think it has been the right move. The velvet rope is what makes Confluence unique, and if we were to focus on growth at all costs, that velvet rope (and probably any loyalty we have built) would disappear.
Did not focus on design and copy for advertisements

In the modern world, it is useless to be creative unless you can also sell what you create. Weโve used advertising to promote other brands as well as our own, and weโve made some mistakes in that process.
Any time you add advertising, you risk diluting the experience for your audience. How much you dilute the experience is based on the quality of your copy, the design of the visual, and the degree to which the ad looks native to the rest of your content. People typically donโt mind being interrupted as long as they are entertained, and weโve based our new advertising strategy on this principle.
If youโre running ads and need inspiration for ads that convert, I highly recommendย checking this out. This article about Morning Brewโs marketing engine is one of the best Iโve read on improving content marketing.
Stuck to content schedule instead of doubling down on what was working
For close to a year and a half, we stuck to one podcast and one newsletter per week. This worked for us, we were getting good marketing analytics, and we didnโt want to break anything that was working. If I could go back in time, I would double down on the strategies that are working in order to double the output.
Over the past few months, weโve done that by sending out newย weekly newsletters for VC jobsย andย deal flow, and itโs instantly tripled our content leverage.
Did not test automations before deploying them

Thatโs myย Zapierย history for the past seven months. I like to automate things, and I hate testing whether things work before turning them live.
There have been a couple of instances where an automation breaks and people get spammed with notifications (if anybody remembers when we announced Commonapp, we had a bug that notified every channel whenever a new company submission wasย created). This obviously hurts your authority, so letting this happen was a big mistake.
Introduced too many pieces of software

This is ourย current software stack. Itโs a mess.
If youโre building software to consolidate the amount of tools that a community leader needs, Iโd love to meet you.
Sold low-ticket subscription offerings
We recently stopped publishing our premium newsletter. This was a $10/month or $100/year subscription for additional content from us. It was a nice additional stream of revenue, but it was never going to scale, and it became a distraction after a while. If we priced this higher, weโd probably keep running it.
Chasing down small subscription checks is a tough business to get into, and Iโd recommend against it.
Under-thinking the legal stuff
Incorporating your first company is pretty scary, and there are a million different things to consider. You should really read intoย what type is right for you, and Iโd recommend getting a second opinion from somebody with a legal background.
We ultimately incorporated as a DE LLC without thinking through the tax implications. The good news is that we can take distributions and only be taxed once at the individual level. The bad news is that it essentially dictated that we were going to remain bootstrapped and not raise funding (which actually worked out in our favor).
If youโre looking to incorporate today, Iโd recommend usingย Stripe Atlasย orย Firstbase. You really canโt go wrong with either, and both have made the process simple.
Did not invest into website until way too late
I created our firstย websiteย in four hours with no web design skills. That was good enough for the first version, but we waited almost two years to get outside to improve it. If you arenโt a web designer, youโre better off finding somebody to help. If youโre building a website on Squarespace, you can find great designers through their partner marketplace.
Wait too late to start tracking traffic through Google Analytics
Google Analytics still confuses me, and we waited way to long to get this set up on our site. We werenโt able to identify traffic patterns, and we werenโt able to focus in on the things that work. Iโve also heard thatย Simple Analyticsย is a good option for those that prefer more privacy.
Did not leverage affiliate programs

People are motivated by money. Weโve been lucky to get social proof and referrals, but we havenโt been able to figure out how to reward people for referring business over to us. Weโre looking at usingย Rewardfulย in the future but still need to think through the logic of how that would work with our existing process.
Selling products, not bundles
All of ourย productsย are individual purchases. Each product has a different purchase price, and each customer has a different journey to purchase.
One of the best pieces of advice Iโve heard recently is to sell bundles, so that itโs easier for people to justify the price. There are someย upselling toolsย available in Stripe, and Iโm still working through how to use these effectively.
If youโre interested in bundling together any of ourย products, let me know, and Iโll see what I can do.
Did not consistently share on platforms (LinkedIn, Twitter, YouTube)
Everybody has a different preference on how they consume information. If you only focus on one platform or medium (thatโs primarily been email for us), you ignore a lot of your potential audience. You should regularly post on all of these platforms, but you should prioritize the ones that you think will result in the most reach.
We think that Twitter is a great place to spread ideas but a terrible place to create pipeline (most crowded platform, and the half-life of content is shrinking by the day). LinkedIn is great if your audience is primarily professionals, but the majority of posts n the platform are self-promotional, and itโs hard for your content to go viral even if you are adding value. YouTube is our favorite platform, and I wish we would have invested more into our channel. Weโre betting YT will be the most popular platform over the next decade.
Relied on the same content structure every week
Even if your content is great, people will get tired of the same structure every single week. You need variety, or people will lose their interest in you. Deep dives like this one allow us to share more of whatโs working / not working, and it keeps the content more fresh and valuable.
Didnโt punish bad actors soon enough
Everybody that joins Confluence is vetted, but we still make mistakes. Some people have come in with the expectation of taking something from the group without giving anything in return. Theyโve spammed channels, bothered members, and hurt the validity of the group. Weโve done an okay job of removing these people after warning them, but I wish we implemented a one-strike policy earlier.