Your developer disappeared

Someone built your software. They were good, or good enough. Then the replies got slower, then they stopped, and now there is an invoice arriving every month from an account nobody can log into and a website you are afraid to touch.

This is the single most common reason people call me, and it is worth saying plainly: it is a recoverable situation almost every time. It does not feel that way from the inside, because the thing you have lost is not really the code. It is the ability to make a decision about your own business without asking someone who is not there.

What is actually broken

Three separate things fail at once, and they get talked about as one problem.

Access. Domain, DNS, hosting, cloud accounts, the app store listings, the error monitoring nobody has looked at in a year, the transactional email provider. These are usually scattered across five to fifteen services, some of them billed to a personal card you have been quietly reimbursing.

Knowledge. How the thing deploys. What the environment variables are. Which of the four database backups is the real one. Why that one cron job must run before the other one. None of this is written down, because it never is.

Momentum. Every change becomes a negotiation. Small requests get deferred because nobody can price the risk, and after a few months of that the software stops being an asset and starts being a thing you route around.

The access problem is the loudest. It is also the easiest. The knowledge problem is the one that actually costs you.

What to do this week, before you hire anyone

Do these in order. None of them require a technical person and all of them make the eventual handover cheaper.

Find the money. Go through the last twelve months of statements and list every recurring charge that could plausibly be software. Vercel, AWS, Heroku, DigitalOcean, Supabase, Twilio, SendGrid, Stripe, an app store developer account, a domain registrar. This list is a map of your system. It is often the most accurate documentation that exists.

Do not cancel anything. I know. Some of those charges are for services that your production database is sitting on, and a declined card can take you offline with no warning and no backup. Cancel later, deliberately, once you know what each one does.

Secure the domain first. Everything else can be rebuilt. A domain you lose control of can be genuinely unrecoverable, and if it lapses it can be bought by someone else within seconds. Find the registrar, start the account recovery process, and get it transferred into an account you own.

Take a backup of anything you can still log into. Export the database, clone the repository, download the assets. Do it now, while access still works, because access sometimes stops working when a card expires.

Write down what you are afraid of. Which part of the system, if it broke on a Tuesday, would cost you the most? That is where the work starts, not with a tidy plan to redo everything.

What I do

I take the whole thing off your desk and become the technical person you actually have.

Recover access. I run the provider recovery processes, consolidate the scattered accounts into ones your business owns, and get the billing onto one invoice instead of eleven. This is unglamorous administrative work and it is usually the first two weeks.

Get it deployable. The functional definition of “you own your software” is that a change can be made and shipped safely. That means a working local setup, a deploy that is one command instead of a ritual, and a backup that has actually been restored at least once rather than merely existing.

Write the runbook. What it is, where it lives, how to deploy it, what breaks, what to do at 2am. Short and real. If I get hit by a bus, the next person reads one document, and you are not back here again.

Then stop the bleeding. Whatever was actually hurting — the broken signup, the payment that silently fails, the page that takes nine seconds — gets fixed once there is a safe way to ship.

Then it just runs. Most people move onto a monthly retainer at this point, because the problem was never really the code. It was not having anyone whose job it is to care whether it works.

What it costs

Recovery work is scoped and fixed-price once I have looked at it, because you have already had enough open-ended arrangements with technical people. A first look is a call and it is free.

If you want someone to keep owning it afterwards, that is a monthly arrangement with no long-term commitment, priced once we both know what the system needs. Set that against a senior engineering hire at $150k plus benefits and recruiting, or against the cost of your software being a thing you are scared of for another year.

When you should not call me

If the software genuinely does not matter to your business any more, do not rescue it. Export the data, take the domain, shut it down, and spend the money on something that makes you money. I will tell you that on the call, and it is a real outcome, not a sales move.

If what you need is a body to work tickets under someone else’s direction, that is staff augmentation and there are firms that do it better and cheaper than I do.

Common questions

Start with a call

Thirty minutes, no charge, and you will leave it knowing what you have and what it will take. Bring the list of recurring charges if you have it.

→ Book a 30-minute call