I've hired someone to build an MVP for a website. The service crawls through and scrapes academics' and journals' websites and notifies users when new articles meeting their pre-specifications are posted.
In theory, once it's up and running, there shouldn't be any need for me to pay a programmer for maintenance/bugs. In practice, I know issues are going to come up. I'm wondering how many programmer hours per month I should expect to need for maintenance alone, on average, assuming I hired a decent programmer. A few hours? Tens of hours? Hundreds of hours?
Thanks in advance!
There a many factors that will affect the ongoing maintenance work:
-how many sites are you scraping and how often are they updated? Every time their layout changes you may need to update your app
-your underlying software (e.g. PHP, MySQL, etc) will likely be updated for security or bugs and you will need to update your app.
-reliability of your app
if everything is pretty steady state, i Would imagine you can get by with just an hour or two per month on avaerage.
I'm starting to compile a collection of resources and tools to help with questions like this at Riffbiz.com. One such article I saw recently discussing the running costs of a SaaS app may be relevant: https://cushionapp.com/running-costs/
Purpose of MVP is to test the water and build upon it as per user acquisition and feedback. Its hard to predict the number but, if website has been developed and hosted with right , you shouldn't need many hours.
Apart from core feature MVP must have detailed analytics, which should give you correct insight of your website.
There are three potential scenarios that will cost you money moving forward:
1. When the structure of the webpages you're scraping changes, you will need to update the scraper. The more websites you scrape from, the more likely this is to be an issue.
2. If the server running the scraper goes down, you will need someone to fix/restart it.
3. When the scraper encounters an unexpected value, you will need to make changes to the scraper. This is common for untested scrapers when they first launch.
Having built and maintained web scrapers in the past, I'd recommend budgeting between 2% and 5% of the original build cost per month for maintenance in the first year.
We generally dial our clients down to 10-ish hours per month after the MVP / prototype launches. In addition to bugs / issues that come up, this also gives enough time to handle admin tasks like backups, etc. Also -- in this phase, software is hardly "done", so we also add / modify features after people start using and playing with the MVP.