Say you manage five brand websites and want to run a survey on multiple websites without building five separate projects. The questions are identical everywhere, but you still need to know which response came from which site.
This is a common setup for companies that run several storefronts, regional pages, or product sites under one brand. The challenge is not the survey itself. It’s keeping the data organized once responses start coming in from different places.
In this blog, we’ll explore how link tagging, segmentation, and cross-site reporting work together, plus the mistakes that cause data to get mixed up.
What does it mean to run a survey on multiple websites?
Running a survey on multiple websites means deploying one identical survey, whether as a popup or a plain link, across several domains while still being able to tell which site each response came from. The survey design stays fixed. Only the source changes.
Most survey tools handle this by adding a small piece of tracking data to the survey link itself. In QuestionPro, this is done through a survey link that can be posted on as many pages as needed, with a hidden variable attached to each version of the link.
Nielsen Norman Group notes that intercept surveys, where a popup appears to visitors at specific points on a site, are one of the most common ways researchers gather feedback during an actual site visit. Running that same intercept survey on several domains just adds a tracking layer on top of a familiar method.
Why track survey data by website instead of combining it blindly
You track survey data by website so you can spot which specific site is driving good or bad feedback, not just what the average visitor across all sites thinks. Without a way to separate responses, a strong-performing site and a struggling one get blended into one misleading average.
This matters most when the sites serve different regions, products, or audiences. A 4.2 average satisfaction score sounds fine until you learn one site is scoring 4.8 and another is scoring 2.9.
Per-site tracking also lets you catch problems early. If one domain suddenly shows a drop in scores, you can investigate that site specifically instead of digging through combined data to find the source.
How to tag survey links using custom variables
You tag survey links using custom variables by appending a variable and value to the end of the survey URL, so each version of the link carries a label identifying where it was posted. A custom variable is a small piece of data attached to the survey URL that gets saved alongside each response.
Every QuestionPro survey has a base URL that looks something like this:
https://www.questionpro.com/akira/TakeSurvey?id=xxx
To tag it by source, add a variable and value pair after the ID. Replace xxx with your own survey ID; these are examples only and are not live links:
https://www.questionpro.com/akira/TakeSurvey?id=xxx&custom1=flowers.com
https://www.questionpro.com/akira/TakeSurvey?id=xxx&custom1=roses.com
Both links open the exact same survey once xxx is replaced with a real ID. The only difference is the value stored in custom1, which tells you the response came from flowers.com or roses.com.
Example: Tagging links for two websites
A skincare brand runs the same product feedback survey on its two storefronts. It creates two link variants:
- flowers.com gets custom1=flowers
- roses.com gets custom1=roses
Each visitor who completes the survey has their response tagged automatically, with no extra question needed to ask which site they came from. Full setup steps are covered in the QuestionPro custom variables help documentation.
How to segment survey results by website
You segment survey results by website by creating a data group for each custom variable value, so the platform filters responses into per-site buckets automatically. This step happens after data collection, in the analysis section of the survey.
In QuestionPro, this is done through system variable grouping:
- Open the data segmentation and filtering section of your survey analytics
- Create a new group where custom1 = flowers.com
- Create a second group where custom1 = roses.com
- Save each group to view its stats independently
You can also combine values. If you run the survey across 40 domains and want to view 10 of them as one combined segment, list those 10 values separated by commas when building the group. This gives you a unit of analysis for that cluster without losing the option to break it down further later.
You compare all sites at once with banner table analysis by setting your custom variable as the pivot question, which generates a single report showing every site’s data side by side against the aggregate. A banner table, also called a pivot table, is a report format that shows response data broken out by one variable across every question in the survey.
To run this in QuestionPro:
- Go to the banner tables analysis screen
- Select your custom variable (for example, the website source) as the pivot
- Generate the report
The output is an Excel file with each site’s results in its own column next to the combined total, so you can scan differences without building separate reports for each domain. Setup details are in the pivot table setup help guide.
Real-world example: Reading feedback from 10 websites at once
A franchise group with 10 regional sites runs one customer satisfaction survey across all of them. Individually, each site gets too few responses per month to draw solid conclusions.
By grouping all 10 sites into a single combined segment while keeping each site’s raw data tagged separately, the team gets a reliable regional sample size for monthly reporting. When one location’s scores dip, they can isolate that single site’s data in seconds because the underlying tagging never changed, only the grouping applied on top of it.
This is the practical value of tagging every response at the source. You decide how to slice the data after the fact, without rebuilding anything.
Common mistakes to avoid when running one survey across websites
The most common mistake is forgetting to change the custom variable value on each site’s link, which causes every response to get the same source tag regardless of where it was actually collected. Always double-check each link before publishing it.
A few other issues to watch for:
- Using special characters (commas, hash signs, spaces) in variable values, which most platforms reject or strip out
- Reusing the same variable name for different purposes across surveys, which makes reporting inconsistent later
- Assuming the base survey link alone tracks source. It doesn’t. Only the appended variable does
QuestionPro allows up to 255 custom variable fields per survey, each storing up to 512 characters, so running dozens of tagged sites at once is rarely a limitation in practice.
Survey on multiple websites vs. separate surveys per site: Which is better?
One tagged survey across multiple websites is better when the questions are identical, and you want combined reporting alongside per-site views. Separate surveys per site make more sense when each location needs different questions, branching logic, or branding.
| Factor | One survey, multiple sites | Separate survey per site |
|---|---|---|
| Setup time | Low, one build | High, one build per site |
| Question consistency | Guaranteed | Depends on manual upkeep |
| Combined reporting | Built in | Requires manual merging |
| Custom questions per site | Not supported | Fully supported |
| Best for | Identical feedback across sites | Sites with different needs |
If your sites share the same customer journey and you mainly want to compare performance across them, tagging one survey is the simpler and more reliable route.
Custom variables vs. UTM parameters: What’s the difference?
Custom variables and UTM parameters both attach tracking data to a URL, but they serve different systems. UTM parameters are read by web analytics platforms like Google Analytics to track marketing campaigns, while custom variables are read by the survey platform itself and stored with the response data.
You can technically use both on the same link. The UTM parameters would tell your analytics tool how someone arrived at the page, while the custom variable would tell your survey platform which site they took the survey on. They don’t conflict, but they also don’t share data automatically, so treat them as two separate tagging systems working in parallel.
Keep the tagging simple as you scale to more sites
The setup described here works the same whether you’re tracking two websites or two hundred. The only thing that changes is the number of tagged links you manage.
A simple naming convention for your custom variable values, matching each site’s domain name exactly, saves the most time as the list grows. It also makes handing off reporting to a teammate far easier, since the logic behind each segment is self-explanatory. For teams building this out on QuestionPro’s survey software, the tagging and reporting tools described above are available on every plan that supports custom variables.
Frequently Asked Questions (FAQs)
Yes, but every response will be anonymous with no way to tell which site it came from. You’d only see combined totals. If per-site reporting matters at all, add a custom variable to each link before publishing it.
There’s no hard cap on the number of tagged sites, only on the number of custom variable fields available, which is 255 per survey in QuestionPro. Most multi-site setups use one variable for the source and never come close to that limit.
No. Respondents see the exact same survey regardless of which tagged link they clicked. The tracking data is invisible in the survey itself and only appears in your response export and reporting.
Yes. You can generate new tagged links for additional sites at any time. Responses collected before the tag was added simply won’t have that source value, so it’s best to tag early.
Yes. The identical custom variable approach works for email sources, print QR codes, or app placements. The variable value just changes to reflect whatever source you’re tracking instead of a website domain.
