Note: The Doctor is out this week, soon to return. This is from 2015 and still extremely relevant, IMO. Enjoy!
I have no idea what you’re thinking. In fact, I can’t even see you. I guess I’m operating in the dark.
The above pretty much describes any web-based business with regard to their relationship to their customers. Unlike a retail store, where you see and interact with your customers face to face (thus lessening the likelihood that you will miss important cues such as a frown or clenched fist), a job board or recruiting site has limited options when it comes to finding out what the customers think. It can: a) rely on anecdotal evidence such as the occasional phone call or conversation at a trade show; b) deduce intention from site behaviors (look! they’re clicking on the new logo! they must love us!); or c) survey its customers.
As you might guess, I vote for c.
Why? Because carefully constructed and regularly administered surveys of both candidates and employers can put you in touch with hundreds or even thousands of customers. The more responses you get, the more likely the data are valid – and the more confidently you can use said data to guide your business decisions.
Are surveys perfect? Of course not. There is response bias (the possibility that only certain subsets of your audience will respond to a survey, thus slanting the results). Poorly worded or constructed questions can produce puzzling or inaccurate results. Low response levels mean you must treat the data with a big shakerful of salt.
But if you are thoughtful and careful about building your survey (or work with a 3rd party that is skilled in such things), you can get truly useful information. I have used surveys for myself and clients in the following areas:
- audience demographics
- your site/service vis-a-vis your competitors
- how your site/service is used by customers
- how well (or poorly) your service is branded
- how you are perceived vs. other alternative services
- customer plans for the future
- product ideas and directions
The subjects are truly limitless. I do advise asking a few standard demographic questions on each survey (no matter the topic) so that you can ensure that this sample will accurately represent your audience. Also, you can’t send out too many surveys – people quit responding. For most sites, 2-3 surveys per year per customer type (candidate, employer) should be plenty. Avoid sending out surveys during the dog days of summer or during the week between Christmas and New Years. Keep your questions to 15 or less. And always test the survey on a few people beforehand to identify any poorly worded or confusing questions.
Don’t operate in the dark. Ask the question. You’ll be pleasantly surprised.[Want to get Job Board Doctor posts via email? Subscribe here.]