Just as finding an employee is time consuming, difficult and really expensive if it turns out the person is bad; the same is the case with software freelancers.
Software freelancers often market themselves as experts, and because of that title, they also charge a significant premium fee compared to employees. The only real economic benefit you get, is you often can throw them out of your company day-to-day, which at least is great if you find out a person isn’t performing.
But we’re getting ahead of ourself. When you find a new software freelancer: how on earth do you know if the person is good?
It really depends on what level you’re seeking the freelancer. If we’re talking outsourcing at 10 USD an hour through online websites, then it’s one answer. If we’re dealing with a specific technology expert at 150 USD an hour, the answer is totally different.
In this post we will look into host testing works in both ends, and then ask the question: can you maybe get a benefit from using the opposite testing methods (so the expensive consultant get same testing as the cheap) [Hint: yes, and it’s a good idea].
How to know if a cheap outsourcing freelancer is good
At this level, you can simply ask for a coding test or see existing software they’ve made.
If it’s a freelancer, you can either:
- Ask for a specific coding test (keep it below 5 hours)
- Pay for a small first task (max. 10 hours). The benefit here is that all will say yes, and it’s a more fair situation.
This works great if you’re technical yourself, as you’re able to figure out if it’s quality work or not.
If you’re not technical a coding test can be more difficult. One option is getting a friend to review the work, another is just trying the project. Without doubt, it’s an extremely good idea to limit the start at max 10-20 hours.
After 10-20 hours, at least something should done. Maybe it’s only backend and you cannot see it visually, but at least, something must be done. This is obviously easier if we’re dealing with mostly frontend or we’re working in a system (webshop, CMS or similar), because most changes are visual quite fast.
However, my most important tip here: either get a coding test and/or keep a limit of 10-20 hours to see if the person really performs.
How do you know if an expensive consultant is good?
If you ask an expensive consultant to do a coding test, or you’re telling them you want to limit the first 10-20 hours – they will most likely laugh, call you unprofessional and leave.
That makes sense. These guys are used to a corporate environment, where there are no questions with these things. They assume professional treatment.
Here you are really relying on their references and CV. Where have they worked before? Were the previous contacts happy with the work delivered?
In addition to the CV, you can also ask if there are any cases or code to showcase. Often will there be something you can review.
The best way, however, is always to talk with a previous customer. If you somehow can get an idea of who the person worked with before, a simple phone call can tell the difference and get an honest response.
Can you mix the testing methods?
As with most things in the world, this isn’t black and white.
When you have a cheap freelancer, you can actually benefit from the same method used on the expensive freelancer. Checking up on their references and calling a previous customer can really help you out. Challenge is it can be much more difficult to get good answers.
When you have an expensive freelancer, you can often ask for some earlier coded projects. Often there will be something public to be shown, which can help you.
But no matter what, just assuming a freelancer is good is a big problem. Because a lot are not. So remember: test them and make sure quality is high.