When you promote your website, you need to know what’s working — and what isn’t.
With Burst Statistics, you can easily track marketing campaigns using UTM parameters and even run A/B tests to find the best-performing version of your campaign. Burst will automatically recognize your campaigns as an A/B test, if you use ‘variation-a’ or ‘variation-b’ in your campaigns. For example, if you have a link in a blog, you could use recipe-blog-variation-a and recipe-blog-variation-b.
This guide explains what campaigns are, how to set them up, and how to run A/B tests in Burst — even if you’ve never done it before.
What is a marketing campaign?
A marketing campaign is any effort you want to track. Examples include:
- An ad on Google or Facebook.
- A newsletter sent to your subscribers.
- A banner on another website.
- Any link that leads visitors to your site.
To track campaigns, you add UTM parameters to your links. Burst reads these parameters and records your results automatically.
What are UTM parameters?
UTM parameters are little pieces of text you add to the end of a link. They tell Burst (and other analytics tools) where your visitors came from.
Example:
https://example.com?utm_campaign=my-summer-sale&utm_source=newsletter&utm_medium=email
In this link:
- utm_campaign =
my-summer-sale(the name of your campaign). - utm_source =
newsletter(where the traffic comes from). - utm_medium =
email(the type of channel).
In Burst you can also use burst_ instead of utm_.
Supported UTM Parameters in Burst
Burst supports the standard parameters:
- utm_campaign – Campaign name.
- utm_source – Where the visitor came from.
- utm_medium – Type of traffic channel.
- utm_content (optional) – Differentiate between two ads or buttons in the same campaign.
- utm_term (optional) – Commonly used for keywords in paid search ads.
Step-by-step: How to set up a campaign in Burst
- Decide on your campaign name (e.g.
my-summer-sale). - Add the UTM parameters to your link using our campaign URL builder or manually. E.g:
https://example.com?utm_campaign=my-summer-sale&utm_source=facebook&utm_medium=cpc - Share this link in your ads, emails, or posts.
- Burst will automatically detect visitors who arrive through that link and record the campaign details in your stats.
Running an A/B test in Burst
An A/B test compares two versions of your campaign to see which performs better.
To set this up in Burst:
- Create two campaign links that are identical except for the campaign name.
- Add
-variation-ato one, and-variation-bto the other.
Example:
Version A:
https://example.com?utm_campaign=my-summer-sale-variation-a&utm_source=google&utm_medium=cpc
Version B:
https://example.com?utm_campaign=my-summer-sale-variation-b&utm_source=google&utm_medium=cpc
- Use these links in your ads. Ideally, half of your audience should see A, the other half B.
- Burst will automatically detect that these belong to the same A/B test.
- In your campaign table, Burst will show:
- ⏳ Hourglass – Not enough data yet to tell. Let the campaign run a while until sufficient visits have come in to decide on a winner.
- 🏆 Gold trophy – Clear winner with significant results.
- ⚖️ Gold scales – A tie; no meaningful difference between the two. If you leave the campaign active, it can find a significant winner, but the difference will probably still be small.
What does a “tie” mean?
If a test shows a tie (⚖), it means that right now there is no meaningful difference between the two versions.
With more visitors (more hits), the test could still eventually produce a clear winner — but if the difference remains small, it is likely that the final result will stay close. In other words, even if one version eventually wins, the improvement may be too small to make a big impact.
A tie is often a good indication that your two versions are performing almost equally well.
Why A/B testing matters
Sometimes two campaigns perform almost the same. With A/B testing, Burst uses statistical analysis to check if the difference is real or just random chance.
This helps you make decisions based on data, not guesswork.
Tip: Start small — test one change at a time, such as a headline or image. That way, you’ll know exactly what caused the improvement.
