The point of these “musings” is just to get my thinking in public; nothing in these is really changing the team’s day-to-day. I’ve upped the cadence of these updates to weekly. If you wait to give updates until everything is clear, you end up waiting forever. This cadence gets people comfortable with ambiguity. It’s always better to continually publish your thinking like this so when you need to make a change, people have come along the ride with you. Communication along the way solves a lot of issues with change management.

We started with a goal of 20 users by EoM. It was arbitrary and prompted the second question of “why users” and what does that really mean. I then took the “what would Reforge do” approach, which led to this process of mapping our product onto the Aha moment framework.

In general, what we’re really trying to do is move our goals up the funnel because you can’t have engaged users if people aren’t signing up and set up. We’re pushing all of our energies to the left on this funnel. We need more signups and then just enough tracking to see that users have gone from signup to set up. We’ll do work on the rest of it later, but doing extra work on anything further down the funnel from the problem is not worth it because you don’t have enough data.

Trying to specify an Aha moment actually made us realize that there wasn’t necessarily an Aha moment and we wanted one. By focusing on it, it does sort of feel like the metric almost created the feature in some ways, but in a useful way because it made us think about how we show our values to users.

The next complexity we encountered was that we do multiple things. You can change your log levels on the fly, but you can also use feature flags. Some people may be coming to the product with different context and may need different Ahas, or there are different things that could prove they’ve developed a habit. There may be multiple Ahas, but maybe there’s a lens of segmentation for people that we can introduce. This actually made us split onboarding so now we can detect which element you’re onboarding for.

Team versus user was another question we faced when defining these metrics. The gut feel here was to do it based on team and then track users separately. Doing it by user felt like it was going to be confusing and inflationary because if one org has 80 people sign up, that’s not the same thing as 80 different orgs signing up.

Jeff Dwyer

I send out weekly Musings to set a norm of thinking in public. In this one, I work through Aha, Setup, and Habit moments for our product.

Hello! Friday Musings (A day early this week because you can’t control 

I can't believe it's just been a week since the closing. Feels like an eternity ago. Happily I feel much more ready to ready to point people to the site than last week.

I'm super excited about the homepage / marketing work that Andrew cranked out. You know how in Monopoly you have to equally build your property, like you can't have one hotel and just a single house on the others? It felt like we'd gotten to 2 houses on the app and onboarding, but we only had 1 house on www and I was feeling unbalanced.

I remain generally jazzed on videos. The lift isn't that big, being able to see the YouTube analytics feels revealing and they feel higher bandwidth than copy. Please give me feedback if you think they're missing the mark. It's tough to be objective in the abstract.

Also feeling good-enough about the state of the inbound funnel. People can contact us. People can signup. We can watch what they do in posthog. ✅ Hard to see an overall picture / numbers etc, but that’s fine for now.

2️⃣-2️⃣ 0️⃣ 2️⃣ 3️⃣

I'm starting to wrap my head around Q2 a bit more. I think the word for April is "V1 Complete". What's that mean? To use the monopoly analogy again, I think it's just filling in a few holes. Holes are things that I feel like we support, or the website says we do, but they aren't 

 real yet. Segments, Log Level Durations, Config with arbitrary clauses, Log Level rates and a real take on Context. Then trying to hold the line there feature wise and spend all extra calories on documentation, client ergonomics, and hopefully continued iteration from onboarding.

The second half of the goal is really just doing whatever we can to fill the pipe with interested users. I'm still leaning towards a Jeff code-free April to help enforce that. At the risk of being too hesitant, I would really love to have a few more users having a happy time before we really make noise. But I think there are good tranche of "semi-friendly" we can mine next week. Start social. LinkedIn in particular. That’s my plan for next week.

I still think [Customer] & [Customer] are huge assets for us and should be our default Priority 1 relative to anything else. The use-case blog posts that we can do with them in a month or two are gold and the learnings along the way are our best feedback loop right now.

So yeah, I need to spend 90% of time outward facing right now. To make that really work, I think that means getting clear alignment on what "V1" means and then letting you all handle shipping that. How much explicit goal setting do we need here?

So far I’d say I’m doing CTO / CPO / CMO / CRO. There’s utility to that in that I can try to move it all forward coherently, but on something like Contexts or GRPC vs HTTP, I’m starting to feel less confident that my instincts are better charting a path there. Do we want me running that discussion like context-FEST? I’m not sure. Don’t need to change now. We’ll chat more about thoughts altogether next week on this.

Then there's the rest of Q2: May and June. It feels pre-mature to make much of a plan here. I could imagine it being: CodeGen, PhoneHome, FeatureFlag must-haves, Config Pointers, Sky Sack... But... I think we need to let the users drive this. If we aren't getting enough incoming flow to do that, then we may need to turn growth into a 4 player game for the second half of the quarter. Like the TMNT of devtools growth.

First blush, I was thinking X Users by end of month. 

This is a bit arbitrary and not really defined

. Also it pushes a second question: “why user?”. We have Users. Teams. Onboarding Complete. WeeklyActiveProjects. WeeklyActiveTeam. It’s also not exactly clear which tool is the place to look.

This looks good to me, so I vote we adopt their nomenclature wholesale.

Setup - Team Project has successfully connected.

Habit - Team makes a Log level change in non Development

These are all denominated in Teams not Users. That seems right to me, but I think Dan had said just use Users so I’m open to thinking about it as users.

Using this to narrow the goals… I think I’d be satisfied with 15 Aha moments this month. And 50 next month.

Executive Update and Goals from Prefab

2023-04-07 Musings

Sign up to see more