• Robert Walker

How Survey Research Can Drive ROAS


Marketing research – notably strategic survey research – has not always provided useful insights for informing a company’s digital strategy. This is especially true for media targeting and ad spending. Why is this?


Well, simply put, insights – even from large-scale survey research studies (e.g., segmentation, or brand equity/perception research) – are typically conducted in a silo, and researchers tend to forget about the notion of linking survey sample sources (i.e., opt-in panels) to available external variables that could be used to make smart media planning decisions. And, because there are relatively few well-known use cases that can act as example, survey research is overlooked as a novel way to build out a digital marketing or media plan.

The utility of survey research would be vastly improved if it could provide the granular linkage needed for more precise targeting. Fortunately, this is changing fast.

Many survey research sample sources are now onboarded with pre-existing segmentation codes from 3rd party providers. This opens up entirely new ways to leverage survey research results – and directly shape targeting and media buying decisions in digital.


Demographic/psychographic coding is nothing new, and the results of pioneers like Claritas’ PRIZM were mixed. They were marginally useful in categories that had clear geo-demographic skews, such as those linked to income or ZIP code (e.g., autos, high-end appliances, etc.).


Be aware that external 3rd-party segmentation codes alone add almost no marginal value. In the case of a well-done segmentation study, distinct segments will likely emerge and 3rd-party codes would be redundant with what can be gleaned from the study itself. But... hang in there with me for another minute.

What does add significant value is how we go about piping the insights from the survey research side to the structured coding on the other side where media decisions must be made.

External 3rd-party codes connect us to the digital world. Once the linkage is built, digital targeting and ad spend opportunities become more clear. For example, Neustar’s E1 segmentation (172 segments, gasp) gives survey research new ways to add value in the digital world. Survey respondents appended with segment codes can be profiled on key survey questions (and vice versa). For example, do we really want to target those highly promotion sensitive people in segments 19, 72, and 113? This helps researchers identify optimal targets for digital and levels of spend. Our survey guidance helps both in targeting (e.g., LiveRamp) and allocating digital inventory (e.g., Trade Desk).

And media choices needn’t be entirely digital – they simply need to be addressable.

Media choices also include addressable TV and can address context. Another coding scheme from RMT assigns motivational profiles to both addressable TV content and individual respondents. Survey research can then combine respondent-level data with motivational profiles, and use that to identify TV shows that are in alignment with their belief systems and personal motivational outlook. Pretty cool.


Importantly, survey research is really the only tool that can effectively assess the issue of creative, which most experts say accounts for up to 80% of the impact of media spend.


For me, personally, survey research can clearly move up in prominence into a more useful advisory role. Survey research becomes a very useful tool in the marketer’s toolkit to drive brand growth and improve efficiency/performance/ROAS. One of the thought leaders in our industry, Joel Rubinson, has more to say here.


CPG/FMCG has had these tools available for a while now – notably retail scanner data, frequent shopper data, first party data, and customized purchase panels. They provide a steady data laboratory for CPG/FMCG experimentation and insights.


But I think the more interesting and exciting opportunity is in non-CPG/FMCG, especially for categories that do not have retail scanning, or 3rd-party verified POS/sales reporting.

Survey-based approaches can absolutely be used to shape digital messaging, targeting, and media planning/buying for non-CPG if done correctly.

So, whew, crazy right? Here are some thoughts/consideration factors as you begin to explore the possibility of utilizing survey research for digital targeting and media spending.

  • If you have only first-party data, onboard external segment codes if you can. This is the most immediate way to ease into the use of targeting methods, and you can begin to experiment and test. Don't waste your time on A/B testing, which is great for refinement. Focus instead on targeting – that's where the gold is!

  • Understand demographic and lifestyle identifiers (e.g., Neustar’s E1 as an example) that may have already been on-boarded by your retail scanning data provider or survey research sample providers.

  • In non-CPG/FMCG, look for opt-in sample source providers with segments already on-boarded, such as Dynata or Numerator.

  • If you have a very small brand, it makes sense to find those few segments allied with your brand. Your small brand may have a large share in key segments: identify segments with higher shares and target them. This can generate vast improvements in ROAS.

  • But don't over-segment! Some segmentation approaches are absurdly huge, such as the popular Neustar E1 scheme (172 segments!). If you can, condense and work with a more modest approach, perhaps 20 or 30, as a first step.

  • For non-CPG/FMCG, work with a reputable researcher and conduct well-designed research with very large sample sizes to get adequate segment representation (remember, the goal is to link backwards).

  • Utilize demonstrably proven survey questions, such as constant sum, replacement vs. addition, share by use occasion, and so on to assess volumetrics. Note that we approach non-CPG/FMCG in the same way as CPG but replace scanning with survey data.

  • Take your high return segments and model them onto the larger universe of individuals aka “look alikes”.

  • Utilize the improvement in ROAS to reinvest in other “look-alike” segments, or in other geographies.


These are new and exciting opportunities! I believe that they will lead to a resurgence of survey research as a new method for optimizing digital targeting and media planning.


Get in touch to discuss further.

© 2016-2020 by Surveys & Forecasts, LLC