The past two posts were mostly about games I am currently playing. It is interesting to see that the World of Warcraft Holy Paladin experience was nerfed twice, and the general Diablo IV play experience is improving. I think there is some subset of grinders who will love Diablo IV when it gets a few more seasons of fixes, and that is all I have to say on the matter—the game is not for me.
I am going to get back to general engineering conversations. I am slightly nervous about doing so because I just realized that I am almost out of topics again. It has been a while since I have found myself in this state, so I will need to invest some thinky time in coming up with new things to complain about.
Today I want to talk about authenticity. Wait, no. I want to talk about authentication. Maybe a little bit of both actually.
Everything about online product authentication sucks. There are so many different ways to authenticate a customer these days, and almost all of them are a trap of some kind. I am ready to go into all of the different ways you can be your own Admiral Akbar, and decide which traps you do and do not want.
The first trap is the support trap. Many third-party auth providers give you libraries to integrate with their platform. The biggest problem with these is when the various Powers-That-Be make changes to their platforms and online stores. If you are natively integrated into an app store, for example, you might find that you will need to make an urgent update to your service or else run the risk of getting shut off for violating some privacy law or becoming a victim of some exploit. It is very frustrating to distract from core feature development for some problem created by someone else out there.
The second trap is the cost trap. Once upon a time, platforms were your friends in finding audiences. Today, they are a piece of expensive marketing machinery. Growth on the platform can sometimes “go viral” if you do it right, but that is a serious case of someone trying to make fetch happen. Many times it comes across as totes cringe. You will spend advertising dollars on these platforms to find customers, and most of the platforms are vicious red oceans at the moment, which implies that the dollar value will be high.
The third trap is the relationship trap. Sure, you will reduce some friction by having an easy button to integrate with Facebooks and Twitters. I suppose it should be Metas and X’s, to avoid a little “well, actually…” The best products I have seen have built deep relationships with their customers—making sure you have a vital path to marketing to your customers is very important. It is a lot of work to ask for people’s email addresses, and most platforms increase the amount of friction for software developers to access this information because they want to keep their audiences inside of a protective moat.
Related to the relationship trap is the knockoff trap. Because you are partnering with a platform for authentication, they will have some bored product manager somewhere who looks at authentication data and if you are sitting on bottled lightning, they might make an argument to compete in your space, either by acquiring a company or building a product from scratch. You can see the Instagram acquisition or the launch of Threads for some examples of this. Apple and Google are also guilty of this with more or less degrees of success.
Using other authentication providers does come with some benefits. It does make your marketing easier to execute, and it does reduce friction for potential customers. It might take some work off of your plate when you are first building out your service, too.
There will come a time when you will have to look at your long-term strategy and decide that having a more direct relationship with your customers is really important. Many of your authentication partners may wish to stay in the middle of that conversation and attach friction or cost to it.
The other issue is that friction is very harmful to new product development. Building relationships with people on the internet is difficult, and leveraging a trusted authentication partner can be a good way to get some early customers into your service.
In the process of writing this article, I searched for the words “consumers” and “users” and replaced both words with customers explicitly. It was interesting how many times each appeared, and how differently this whole post reads when thinking about customers.
Many of the platforms today that authenticate customers are in the marketing and advertising business and have leveraged colossal amounts of venture funding to buy their way into everyone’s heads. They live there mostly rent-free for customers. Someone somewhere will have to pay for that and unfortunately, that is you and me.
This is one of the reasons that Tim Sweeny is my Games Industry Person Of The Year, every single year. He has done a phenomenal job of ensuring Epic Games has deep relationships with developers and players alike and built a great emerging ecosystem. I also have some mild ideas about how web3 will become more meaningful for long-term player identity, but there are other problems that web3 will solve first.
Writing this post was a struggle. I am conflicted about the ease of leveraging high-scale platforms to jumpstart a customer base because it can “cheapen” your relationship with your customer, and also it erodes a sustainable competitive advantage. I think there is a time after you have found product-market fit when leveraging platform partners and third-party authentication makes sense, and I think that it is something that needs to be done thoughtfully. Most people just dump a generic auth provider in front of their potential customers and hope for accidental long-term success. I am sorry if I offended you with this statement, and in the immortal wisdom of one of my favorite eighties musicians… maybe you needed to be offended.
I tried to create a link to the artist and the album in question to make zero million dollars in affiliate fees, and I was told that there is an error. Maybe my time as Amazon’s worst affiliate marketer has come to an end. I had a good run. I have zero regerts.
I will see you all next week!
