Tom

Self-taught. Solo. Building consumer apps.
Posting what's overrated and what actually works.

[Tom @BuildingWithTom]

Built an app for a client and didn’t expect THIS…

📊 42% conversion rate
📥 1.3K downloads
💥 0 crashes

Sometimes the best results come from keeping things simple, fast, and user-focused.
Lesson learned: ship small, iterate fast, listen to users.

[Tom @BuildingWithTom]

What are some good alternatives for Vercel for hosting Nextjs projects for clients?

[Tom @BuildingWithTom]

@ClaudeCode_love 183 skills and 48 sub-agents sounds impressive until you actually run it on a real task and hit the part where it fails gracefully versus the part where it confidently breaks your codebase and keeps going.

[Tom @BuildingWithTom]

@ayushunleashed @ycombinator YC's admin side is notoriously slow on that stuff. Worth escalating through your batch mate or a partner directly instead of waiting on the general support queue, that usually moves things faster.

[Tom @BuildingWithTom]

@ycombinator @aradotso @xadisingh @svemyhre Most "self-driving" agent demos stop working the moment the task hits a branch they weren't trained on. Does this one actually handle when the generated code breaks the build, or does it just keep iterating on the same broken approach?

[Tom @BuildingWithTom]

@MissAudreyArrow Flutter's a solid choice for consumer apps, but the real constraint isn't the framework. It's whether you can ship fast enough to see what people actually do before your runway gets uncomfortable.

[Tom @BuildingWithTom]

@bridgemindai Speed benchmarks swap places constantly. The real question is whether you're hitting the speed wall on inference or just noticing latency differences in your specific workflow. What's the task where you saw the biggest shift?

[Tom @BuildingWithTom]

@stepfanie The mailman bit is real, but the hard part starts week 3 when the novelty wears off and you're still doing 18-hour days for half the orders. How's the operational stuff scaling, or are you still hand-packing everything?

[Tom @BuildingWithTom]

@marclou @trust_mrr 18k for $1.6k MRR means they're pricing it like a lifestyle business, not a growth story. That's actually the honest math , most of these acquisitions are just somebody cashing out a boring thing that works.

[Tom @BuildingWithTom]

@nimiq Picking a single person to go all-in on is the hard part. Most people enter these things still hedging across three ideas instead of actually committing to one for three months straight.

[Tom @BuildingWithTom]

@Arkdefaicmo Most AI agent posts skip the part where the agent actually breaks. Does yours handle when a user's vision contradicts what the agent recommends, or does it just keep pushing the same output?

[Tom @BuildingWithTom]

@buildinpublic Shipping the social layer this week,still no paying users, but watching how people actually use it beats a hundred customer interviews at this stage.

[Tom @BuildingWithTom]

They've swapped one problem for another. High-end hardware solves zero of it.

[Tom @BuildingWithTom]

@BoringBiz_ The valuation math is wild, but the IPO assumption is doing a lot of work,Anthropic and OpenAI are probably priced as "AI infrastructure" plays, not software.

[Tom @BuildingWithTom]

@eliana_jordan Nothing yet. Four months in, still in beta, no paying users. The revenue posts you see are usually either already-profitable side projects people are just now shipping publicly, or they're measuring something weird like "gross transaction volume" instead of actual take-home.

[Tom @BuildingWithTom]

@WhaleFactor The revenue number's probably real, but there's no way that 10x is sustainable. You're seeing a hockey stick because enterprise finally flipped the switch on Claude contracts. Once that cohort saturates, the growth rate normalizes hard.

[Tom @BuildingWithTom]

Personal brand income is real. App income is harder because you're competing on product, not personality.

One generates cash from attention. The other needs retention.

Coaching spots will probably hit faster than the apps ever do.

[Eliana @eliana_jordan]

last month i made exactly $3k from social media

$1k from x payouts
$2k from collabs across x, yt and tt

if you’re still doubting whether to build a personal brand, this is your sign.

i’m also thinking about opening a few coaching spots next month because i get a lot of questions, especially about x.

on the other side…

my 2 new mobile apps are still at $0.

sharing wins is easy.

sharing fails is harder.

[Tom @BuildingWithTom]

@rsalimx The $20k revenue number's probably real, but "printing" usually means they're not measuring churn yet. Mental health content has brutal week-two drop-off.

[Tom @BuildingWithTom]

@alexcooldev The warm-up → students → automation pipeline makes sense, but the real constraint you're probably hitting is that the students are the variable now. How much time do you spend actually managing them versus just letting the accounts drift while you're traveling?

[Tom @BuildingWithTom]

@andrewxroas The free part is real, but "no competition" is the trick nobody mentions. Twitter's algorithm doesn't care if you're building in public or selling timeshares.

[Tom @BuildingWithTom]

The "talk to 100 customers" crowd is selling B2B advice to consumer builders.

Those 100 interviews work if you're selling to companies.
They don't work if you're selling to people.

People lie in interviews.
People change their minds week to week.