Integrated Consumer Panel.

Glow's Integrated Consumer Panel allows you to DIY capture 100's to 1,000's of responses to your questions. To create one,

Go to the Capture tab.
Click Create Channel and select Integrated Consumer Panel.
Give your channel a Name (optional).
Pick your Country - default is Australia but pick from 47 countries. 1 country max per channel.
1. Pick a Language where applicable. Only major languages available. 1 per channel.
Set your Response Limit - default is 1000, update as needed.
Adjust your Days in field - default is 3 days, but this can be adjusted depending on your sample size, e.g. 1-2 days is feasible if you only need a few hundred.
Prevent duplicates using either:
1. Exclude respondents from - select channel(s) to exclude, only use if multiple panels. OR
2. Enable IP protection - prevents duplicates based on respondent's IP. You can turn this on always if you like.

Incidence Rates

If your survey has exit logic, e.g. then an Incidence rate field will appear. Incidence rate determines the cost of your survey, and is the percentage of your respondents you estimate will qualify for your survey. The formula is IR = Completes / (Completes + Exits). So, if you capture 500 completes and 500 exits, then your incidence rate = 500 / (500 + 500) = 50%.

To get the "genpop" IR, you need to divide your target audience by the number of adults in your market. In Australia, that's 20 million. In NSW, that's 8 million.

XX million / 20 million adults = XX% IR

If you know the denominator, i.e. 20 million adults AND
If you know the numerator, i.e. your target audience,
Then you know the Incidence rate!

Examples:

320k FTE teachers to 515k registered teachers / 20 million adults = 1.6% to 2.6% IR in Australia.
Uni students = 1.6 million / 20 million adults = 8% Genpop IR.

Here's a list of example incidence rates in Australia:

Own home: ~66% (2/3)
- Mortgage ~33% (1/3)
- Outright ~33% (1/3)
Rent: ~33% (1/3)
Have credit card: ~50% (1/2)
Have transaction account: 95%+
Drink alcohol: ~75%
Have pets: ~60%
- Have cats: ~30%

Google, AI, and your own previous research and knowledge are your best friends here!

Researchers in particular should know a lot of incidence rates off by heart. ♥️

The "genpop" IR can only be improved with custom targeting, but targeting outside of age, gender, and location (e.g. targeting homeowners vs. renters or targeting parents with kids 0-18 etc.) is only available for Managed Services, not via DIY channels.

Channel Cost

Our cost model is:

Micro LOI (0-2 min): $1.95 AUD/ £0.95 GBP/ $1.45 USD + 1 hosting credit p/response
Mid LOI (2-5 min): $2.45 AUD/ £1.25 GBP/ $1.85 USD + 1 hosting credit p/response
Long LOI (5-10 min): $3.95 AUD/ £1.95 GBP/ $2.95 USD + 1 hosting credit p/response
Max LOI (10-15 min): $5.45 AUD/ £2.95 GBP/ $3.95 USD + 1 hosting credit p/response
Exits / Over Quotas: $1.00 AUD/ £0.50 GBP/ $0.70 USD + 0 hosting credits p/exit or quota

For n=500 completes, at 50% est. incidence rate (500 est. exits), at 4 min LOI (Mid) in AUD,

The cost is:

500 completes x $2.45 AUD = $1,225 AUD + 500 credits
500 exits x $1.00 AUD = $500 AUD + 0 credits
Total = $1,725 AUD + 500 credits

See article: Plan & Hosting for more info on hosting credits.

Stop channel when cost is reached

Your channel will continue running until your Response Limit is reached, not when your Estimated cost is reached. So the Actual cost can vary from the Estimated cost due to a different Incidence Rate (IR) in field.

To prevent going over budget, tick Stop channel when cost is reached and input your max channel budget. This will stop the channel when your budget (e.g. $1725) is reached instead of when response limit is reached (n=500).

Targeting

Click Targeting and set your demographic targets. Default is set to nationally representative by gender, age, and location. If you want to change targets, go for it - just ensure your target numbers add up to your Response Limit. If you make a mistake, just select the relevant Preset to revert back.

Targets are not hard quotas, they're soft. The platform will do its best to reach your targets but they may be over and under. For hard quotas, use Tag Quotas.
Targeting is covered in more detail in this article: Targeting.

If Targeting is not turned on, then anyone can enter your survey without any sample controls, so will skew towards the fastest responding demographics. Both Targeting (soft) & Quotas (hard) are therefore needed to get a balanced sample.

To target outside of age, gender, or location, then get in touch with us to discuss your target audience. We have over 200 custom targeting variables to pick from, e.g. parents of kids 0-18.

How many Responses do I need?

n=1000 will get you a ~3% margin of error, which is what most researchers want, great for full-service research, published research, or presented research. Robust enough to crosstab.
n=300 will get you a ~5% margin of error, great for quick turn research or research on a budget but not as robust to crosstab due to lower statistical reliability. AND
n=100 will get you a ~10% margin of error, which is the lowest accuracy researchers will accept and also great for soft launches but virtually no ability to crosstab reliably.

So, pick any number from 100 to 1,000, OR

More in the multiple 1,000’s (e.g. n=3,000) if you have a LOT of interlocked segments and brands you want to cut your data by and you need a statistically robust sample size for each one. E.g.
- Non Interlocked
  - Males = 1500 (50%),
  - 18-34s = 1000 (33%), and
  - Sydney = 600 (20%) but
- (Triple) Interlocked:
  - 18-34 Sydney Males = 100 (still statistically robust)

When your channel is ready, click Save, then click Start. Responses will start arriving within minutes / hours in your Account.

What happens if I click Stop?

This will stop the Consumer Panel Channel and prevent further responses from arriving in your account. The Consumer Panel cannot be restarted again so please create a new Consumer Panel channel if you still want to capture more consumer responses.

Covered in separate Help Scout articles

Within the Advanced fold: