Comment by TuringNYC | Hacker News Reader

Comment by TuringNYC | Hacker News Reader

Parent Full thread

+10 on point 3. Even as a highly technical CTO, it is shocking what outsourced dev agencies will try to do. I've seen dev agencies try to conclude engagements with a "successfully working app" but havent delivered any source code, or when source code is delivered it cannot actually be compiled into the successfully working app.

As a technical CTO I can smell this a mile away and work it into SoWs, but I wonder how many non-technical co-founders fall into this trap.

When you need updates or Phase II, now the rates go up massively. No one else can change it and the outsourced dev agency now has a monopoly. Usually the non-technical CEO/founder will now have to find a new agency, who might pull the same trick or some variation of it. It can be subtle:

- You have no code

- You have code, but it wont compile

- You have code, it will compile, but not into the version you're declared success with

- You have code to the client/app, but not to the backend

It makes me shudder to think people actually hire dev agencies to build their early stage product. Having worked with some in the past in a different life I absolutely despised those people for being both terrible engineers and people (my experience working with big outsourcing companies).

Why hire people who don’t care that much to build something the founders definitely care a lot about?

Are there any examples of this working out or are these the 90% of startups that fail?

Personally I’m very technical and I think the best for me would be a partner who is non technical to help me with marketing and sales (something I have no experience in). Would that sort of duo be the best? I can build whatever mobile web backend AI algo shit. That’s easy for me. For my partner I would prefer they are the same level of competence in their side of things.

One of the reasons that I am working on the project that I'm on now (for the last couple of years), is because a friend of mine had set up a 501(c)(3), and was looking to develop an app, out of his own pockets, and through solicited donations.

The main problem was, he didn't really know what he wanted (non-technical).

He asked me to sit in on a couple of his enquiries to developer agencies.

It was a shitshow.

I had respect for the one chap that basically said "Get back to me, when you know exactly what you want."

I had absolutely zero respect for the agency that said they would deliver a full MVP, in three months.

After that, I just said "Screw this. You'll just get ripped off. I was looking for something to work on, so it might as well be this."

It's been a long process, but we are now in the home stretch. I don't even want to think about how badly he would have been screwed, and I'm quite un-thrilled with the ethics of the folks that claimed they could deliver the app in three months (fairly large-scale backend, and high-quality frontends for mobile clients).

This speaks to me. Around 2012 or so I had a semi-technical friend (PM, could tell when there was a lot of bullshit being thrown around but not a dev by any stretch) with an idea and he needed a Facebook page/app. Remember this was back when people were building a lot of brand experiences inside of their company's FB pages, and this was similar. I quoted about 30-40k. He got a quote from one of these offshore places for about half that and went with them. 60k in billings and change orders later, he comes to me asking for help. It's an absolute shitshow. Code was terrible, absolutely no organization, no error handling, no logging, obviously not a single test anywhere. It was a pretty stereotypical example of offshore devs trying to do as little as possible to show "progress" while billing day rates because everything now was considered a change. Multi-page emails listing out problems and asking for status on 15 different tickets would get a response 23 hours later of "will look into this quickly thank you :)", etc.

I offered to rewrite the entire thing from scratch for a flat rate 15k and some equity but was turned down. Last I heard they were about 100k in by the time it "launched" and it shuttered maybe 6 months later. So this guy burned 100k but more importantly about 2.5 years of his life on this thing while a bunch of terrible devs and a charlatan CEO made off like bandits.

> "Get back to me, when you know exactly what you want."

This is so important if you want to maintain a good relationship. I'm working with a client who has very little money and while the system he wants is almost trivially simple, he's funding it out of his own pocket. After he initially was bouncing from "cool idea" to "cool idea" and had a heart attack at my very reasonable price, I told him to take a step back and write down the most important features and I'd quote that.

He did that, I built it and he loves it. Now he wants version 2 and we're back to the "wouldn't it be great if..." all over again. Sure, it would, but you already said you can't afford all that crap. Now I'm back to the "stop. Think about what you really need, forget the cool IoT features and focus on the basics then get back to me."

Sorry, needed to vent :-)

You’re a good person that’s awesome. Sometimes it’s necessary to guide people to the water (but ultimately it’s their choice what to do).

Nice good luck to u and ur friend

Knowing exactly what you want is really important, and where most non-tech people fall down.

It's very easy to come up with vague, hand-wavy "big pictures," and a different matter, entirely, to make it happen. Those "details" will kill you.

That's why having not just a technical founder, but one that has experience shipping is important (IMNSHO).

Sounds more like you got suckered into working for cheap by someone you know. Unless the app is drastically different than the last CRUD app the firm did, 3 months is plenty of time to run find-replace on a codebase.

Nope.

It probably would have been quite different, and I guarantee that they would have come up with some bolierplate crap that they had their offshore devs whip up for next to nothing.

The demographic we're Serving has some very particular needs, they don't have much money (501c3 means nonprofit, if you are not familiar with the designation), but there's tremendous sensitivity in how the data is handled, and also in how the UI is presented. Small features will make (or break) the app.

Long story, and not one I'm particularly interested in sharing, at the moment.

It's all good.

BTW: I'm not working for "cheap." I'm working for free. I'm enjoying it.

I read it he volunteered to work for cheap for a friend, out of charity, or pity even. Your second sentence I don't quite understand, you mean the original shop did a bad job? Well, yes.

>> Why hire people who don’t care that much to build something the founders definitely care a lot about?

Because people dismiss engineers as low-value, until they painfully realize otherwise. It is the same reason Google/Facebook/Stripe did so well compared to other companies -- because Google/Facebook/Stripe realized the strategic value of engineers and compensated them well and aligned with long-term interests. Other companies cut Engineers and constantly squeezed them until they realized (too late) that they had no strategic power left. The stock market over the past two decades is a fantastic A/B test on people strategy and success.

To be realistic, there is one good reason to outsource dev -- it is to very quickly get a demo-able MVP so you can get funding. You have to recognize that the demo-able MVP is most likely throw-away and you have to be ready to re-build from scratch once you get funding for your own team.

I work at a startup that outsourced its early dev, somehow got sold to a large company.

We had a technical CTO that really wrangled everyone, but he left not soon after we got bought out, dev has been a shitshow ever since.

I've consulted in this sort of role for a startup which hired a dev agency. A few months in, we were behind schedule. I've asked them access to their git repo and built some graphs with the commits and LOC changes of every dev we paid for. It became clear that these devs were working on multiple other assignments and were not really dedicated to us. Without these sort of insights, it's hard to judge if they are honest or not.

I'll make suggestions to anyone doing this:

- For slightly more legal protection, the code should be in YOUR OWN git repo, not the dev agency's repo

- If you are giving some upfront $ or milestone payment, you should also contractually now own the code up to that point in time

- You should not declare a milestone success (and should not pay) until you can independently compile the app straight from git (front end and backend) (Also, watch out for hard-coded domains where you are secretly pointing to their backend and not to your independently compiled one)

- App QA should ONLY be on the version of the app where you compiled/deployed the full stack yourself

- You should deploy the app into your own cloud to ensure you actually have all the pieces. Super extra credit: Build Terraform/Ansible/Chef into the SoW and final payment should depend on deployment into your own environment by YOU (not the agency)

- You should learn how to build/compile/deploy yourself or hire an independent person (not the devs) who can do this for you

If you can do all this, you're either already a developer and don't need to outsource it, or you have the ability to learn how.

In general, I agree. There is enormous hidden equity in being a technical co-founder that is unappreciated until non-technical founders are burnt several times.

That said, I did present two ideas for those who want to venture into this:

- You recognize that the "success" you have in the working app isnt actually a success and plan accordingly

- You hire someone independent who keeps the third parties honest

Another one:

- You cut corners, get a working demo, raise good VC funds and hire a well-compensated technical co-founder or founding engineer who is compensated sufficiently to be happy short term and has sufficient equity to want long term success

...or don't have time

Oh yes, there's so many things that you should do. This was only to grasp the code they were doing. You need a technical advisor who is on your side and can challenge the dev agency. You also need to strike a balance between giving them some slack and fighting to get a solid solution. We relied on the dev agency until we got the initial 2mln investment, afterwards we hired internal devs who took over. But if we scared them too much we might have compromised the investment.

Excellent but dangerous advice. I really hope non-technical people don't read the tips here and think they're now prepared to navigate this process.

100% agree. I've yet to see a startup succeed w/o a technical co-founder. I do angel investments and have never cut a check to a founding team w/o checking for the above checklist. For teams with good founding CTOs, they get it for free if the person is sharp about process.

But the whole discussion was what non-technical founders would do, and this is probably the second-best advice I could give. Moreso, to recognize the myriad risks and understand things arent quite as easy as posting a project on Upwork.

The first-best advice I could give is to get a technical co-founder.

> I've seen dev agencies try to conclude engagements with a "successfully working app" but havent delivered any source code, or when source code is delivered it cannot actually be compiled into the successfully working app.

The recliner I bought 2 years ago has an iOS app that basically works, but has many flaws. It's clear the company does not have the source code. The developer, in the app store, has a name with Chinese characters and the company is based in the US. It was right after the pandemic, that I bought it, and complained to their support. I was still in the window of returning it for a full refund, so they told me "we'll get right on that" and fix the problem I reports. Almost 2 years later and not a single update to the app.

EDIT: company is Human Touch, chair is the Gravis ZG.

EDIT2: sorry, it was the Copyright message that was in Chinese. https://apps.apple.com/us/app/gravis-chair/id1435036207

Last update Apr 2020.

I had one of their recliners. The leather got pretty bad. Now I have the Altwork station though I no longer work 27/7.

Reply on news.ycombinator.com