Search

An 'omnibus' survey? Huh?

Whenever we put out information about our omnibus I get asked the same question: “What is an omnibus survey?”… And since I have a little blog on this website to talk about strategy, messaging, and polling, I feel like this may be a good avenue to describe what an omnibus is, and the pros and cons of participating in one.


When a client wants to ‘go into the field’ with a survey, there are a lot of things that need to happen. Let’s say you want to know ‘what Americans think about x’… Well, we need to make sure we are talking to said Americans, and that our response group is representative of the population. This requires pollsters to ask demographic questions at the beginning of each survey… and a lot of them… Gender, age, race, political affiliation, geography, etc.


These questions need to be asked every time, no matter what kind of population we want to survey … We may not want to only talk to 95-year-old white ladies from Cedar Rapids… Or we may! Either way, we have to ask to make sure we know who we’re talking with, and this takes up time (read: money).


But before we can even field the survey, we need to do nerdy things like draw subquotas of populations, program the survey into the system we’re using (call center, website, text link, etc.), define the banner specs so we can translate raw data files into readable tables, and look at data universes to ensure we’re able to find enough people who will answer our questions.


All of this ‘behind-the-scenes’ work has to be done for every survey, regardless of how long the survey is. And it’s time consuming, and thus, costs money. Problem is, what if you only want to know if your target group favored vanilla, chocolate, or another flavor of ice cream? Well, you would still have to go through all the rigmarole to set up a survey to ask your one question. And that’s expensive. What’s not expensive is to add additional questions on to a survey that is already set up. Thus comes into play: the omnibus.


So to answer ‘what is an omnibus survey’, it’s a survey that has all the behind-the-scenes work taken care of, and then goes into the field with different groups wanting to ask one or two or even five questions – because, again, once it’s set up, asking a few additional questions isn’t as large of a lift as getting the survey into the field the proper way in the first place.


So, let’s say a pollster has 5 clients who all have one question they want to ask Americans, or voters, or whomever. That pollster may do an omnibus survey, where they prepare one survey, include each of their clients’ questions, which oftentimes are on completely different topics, and then provide each client with their own set of data on their questions in addition to all the demographic information asked of respondents. So each client essentially gets to have their own poll, for a portion of a typical survey’s cost.


Sounds great, and it works well for a lot of scenarios. But of course, there are sometimes an omnibus isn’t a great fit for your needs. For example, if you have a LOT of questions you’d like to ask, it may be more cost-effective to field your own survey rather than paying the otherwise-affordable per question cost.


Further, the demographics are fairly straightforward in omnibus polls, meaning they’re generally nationwide, representative of the general population… Well, if you were wanting to know if someone could live without an industrial strength hair dryer, you may only want to do a survey among Druidian princesses, and thus, the omnibus isn’t for you, sir.

However, if asking only a few questions to a large population is your bag, I would suggest an omnibus… Speaking of which, I know of a company who has their quarterly omnibus going into the field soon (depending on when you’re reading this)…