When I’ve written job postings, there have been a few times when I’ve asked myself, “Is that the right word?” or “Will candidates outside the company understand this?” It would be great to find out answers. Well, we can by using a technique called A/B testing.
In short, A/B testing is when you create two messages that are exactly the same except for one factor and then measure which message performs better. It’s important to note a couple of things about this definition:
- Keeping the distinction between one-factor. If there’s only one factor difference, then you know what candidates are responding to – the one factor that was different. If we change two or three factors, then it would take several iterations to figure out the best combination of elements. That could be a challenge from both a time and resources perspective.
- Understanding what “better” means. Let’s say that we conduct A/B testing on a job posting. In Group A, we receive a lot of applications, but many are unqualified for the position. In Group B, we receive a small group of applications, but a vast majority are qualified for the role. If the goal was quantity, then Group A achieved the result. If the goal was quality, then it’s Group B.
One of the reasons I’ve been thinking about A/B Testing lately is I wonder if organizations are using this technique to refine their sourcing strategy. I could see it making sense to run A/B Testing with two different job titles. Especially if you work for a company or industry that uses unique titles. For example, if you’re recruiting for an accounting ninja, is that the same as a payroll clerk? Or if you’re hiring an administrative assistant III, would that be comparable to an executive assistant?
Titles aren’t the only type of A/B Testing. It could make sense to examine the benefits being promoted or the company values being marketed in recruitment messaging. What is the right combination of words or phrases that gives you the result you’re looking for?
If you haven’t considered A/B Testing, now is the perfect time to start testing it out. Create small experiments and evaluate results. Slowly your recruitment messaging will come into focus. That way, you don’t have to wonder – like I did – if you’re messaging is on target.
Image captured by Sharlyn Lauby after speaking at the Society for Human Resources Annual Conference in Washington, DC9
Iain @ ISL says
I’m a huge convert to this A/B testing thing. Once you’ve mastered it with stuff like job postings, try split testing on landing pages, homepages, everywhere…
The site back with your Google Analytics and have some fun. It can be a mind bender, but it can also prove to be a really useful tool for optimising content.