nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

How do you track traffic and build a proper marketing funnel with Atlassian’s attribution gaps?

Marie -Teamlead-

Hi fellow marketers! 👋

I have a question that’s been keeping me up at night, and I’d love to know if anyone has found a clever workaround or if we’re all in the same boat here.

We’re trying to build a clean marketing funnel for our apps, specifically tracking initial visits to our marketplace listings. But as we all know, Atlassian tends to strip away a lot of referrer data and paths, making clean attribution incredibly difficult.

When trying to connect the dots between external channels (like organic content or community posts) and actual marketplace views/evaluations, the data gets super blurry.

How are you guys solving this?

Do you rely strictly on custom UTM parameters and just accept the dark traffic gaps?
Have you managed to build reliable funnels using third-party tools or Atlassian Vendor reports?
Or maybe there’s some obvious native feature or trick that I’m completely missing?

Would love to hear your honest experiences, pain points, or any setups you’re proud of. Let’s brainstorm!

photo_2025-09-02 12.28.32.jpeg

Comments

Mariem Daghbouji _Caelor_

Hey! 👋 Honestly, this is something I've been curious about too, and I'd love to see what other vendors say in this thread. Here's our setup, for what it's worth:

For everything coming from external channels, we lean heavily on UTM links and Google Analytics. Any time we point traffic at our Marketplace listings (from our own website pages and blog posts, our YouTube videos, community posts, etc...), we tag the link with UTM parameters so we can see in GA which channel and which specific piece of content is actually driving clicks, and roughly where those visitors land (the listing, the trial flow, the support section, and so on).

Then for what's happening inside the Marketplace itself, we analyze the Atlassian Marketplace Search Keywords report. That helps us understand what people are actually searching for when they find us, which is useful for drafting our listing copy and keywords.

That said, I'm in the same boat as you on the blurry middle. Connecting an external touchpoint to an actual evaluation/install is still fuzzy for us. Really curious if anyone's cracked it better with other native or third-party tools! 😄

Natalia_Kovalchuk_SaaSJet_

Hi Marie -Teamlead-!

I completely share your frustration. 😊

Today, traffic is no longer one of our main performance indicators because the numbers often don't reflect the reality.

We consistently use UTM parameters across all our external channels, like community articles, social media, partner campaigns, and newsletters, to understand which content and campaigns are driving engagement.

For reporting and analysis, we combine data from Atlassian Marketplace with Looker Studio dashboards to monitor:

- Marketplace visits
- Trials
- Trial-to-subscription conversions
- Active subscriptions
- Churn rate

The challenge is that traffic and even traffic-to-trial conversion metrics can be misleading. Attribution gaps, multiple touchpoints, and limited referrer visibility all contribute to this problem.

As a result, we've shifted our focus toward business outcomes instead of traffic metrics:

- Trials
- Paid subscriptions
- Trial-to-paid conversion rate
- Churn rate

These are the performance indicators we can trust the most.

HEMANT SAINI

Hi @Marie -Teamlead-

You are not alone. The attribution gap is real and it keeps a lot of us up too.

Atlassian Marketplace strips most referrer data on purpose. When a user clicks from your site or a community post to marketplace.atlassian.com, the referrer is dropped. UTM params also get stripped if the user navigates around the Marketplace before hitting your listing. So the funnel goes dark right at the most important step.

Here is what we do today. It is not perfect but it gives us about 70 percent visibility:

1. UTMs plus a landing page bridge
We never link direct to the Marketplace listing. We send all traffic to a page on our domain like /products/app-name/marketplace. That page has one job: fire analytics, then auto redirect to the Marketplace listing with a fresh UTM. You still lose people who browse the Marketplace first, but you capture every click you control. We use a 0 second JS redirect so the user barely notices.

2. Atlassian Vendor reports plus your own analytics
We pull the Vendor Installs and Views report daily and match it against our Google Analytics or Plausible data. Match by date and geo and you can infer a lot. Example: We post in Community on Tuesday, see 80 visits to our bridge page from community.atlassian.com, then see a spike of 60 Marketplace views from the same country on Wednesday. Not exact, but directional.

3. Unique coupon codes per channel
For paid campaigns we create private listing links with promo codes. The code is tied to the campaign. When someone installs with that code, we know the source. Works for ads and newsletters. Does not help for organic.

4. Post install survey
We added one question to our onboarding: How did you hear about us. Options are Marketplace search, Blog post, Community, Peer, Google, Other. It is self reported and biased, but over hundreds of installs the pattern is useful. We track it in our analytics.

5. Third party tools
Tools like Appfire Analytics, ForgeMetrics, or SixFive can help, but they still sit behind the same Atlassian wall. They are better at in app behavior after install than top of funnel. For the click to view gap, there is no magic tool yet because Atlassian does not expose that clickstream.

What we accept as dark traffic
Marketplace search, category browsing, and Related apps traffic. We treat that as a bucket called Atlassian Organic. We optimize listings for it, but we do not try to attribute it to external work.

Native trick you might be missing
In Vendor reports there is a field called Referrer Domain for views. It is often empty, but when it does populate it is gold. We export weekly and pivot on it. Sometimes you catch dev.to or reddit or a partner site that you did not expect.

Honest pain points
1. Cannot do true multi touch attribution. If a user reads a blog, sees a Community post, then installs a month later from Marketplace search, we lose the path.
2. No webhook for listing view. Installs have webhooks, views do not.
3. UTMs do not persist. User clicks your UTM link, then browses two other apps, then comes back to yours. UTM is gone.

Setup we are proud of
We built a simple funnel in Looker Studio:
External click to bridge page
Bridge page to Marketplace outbound
Vendor report views by day
Installs by day
Trials to paid
We cannot link rows 1 to 3 for an individual user, but the cohort view by week is close enough to see which campaigns drive lifts.

Would love to hear if anyone has cracked the Marketplace search attribution. We still call that dark traffic and just live with it.

What is your current stack for tracking

Marie -Teamlead-

Hi Mariem!
Thanks so much for sharing your setup.

To be honest, even just hearing that you are facing the exact same "blurry middle" is strangely comforting! It’s good to know we aren't the only ones scratching our heads over this.

If I understand correctly, you still tag absolutely every external link pointing to the Marketplace, even though GA4 likely flattens a lot of that data into generic referrals or direct traffic once they hit the listing, right?

I’ve been wondering if the Search Keywords report could somehow help bridge this gap - like using keyword search spikes to directionally track if an external campaign actually triggered more intent. Do you think there's any logic to that, or is it better to just kфeep keywords as a standalone pool for copywriting?

Thanks again for jumping in, love how active this thread is turning out to be

Marie -Teamlead-

Hi Natalia!
Thank you so much for your reply, it’s really valuable.

I completely agree with your strategy. Shifting the focus to hard business outcomes like trials, subscriptions, and churn is a very strong approach. Traffic numbers can be so misleading when the attribution is broken, so building your analysis around metrics you can actually trust makes perfect sense.

Thanks again for sharing this, it gives me a lot to think about!

Marie -Teamlead-

Hi Hemant!
Wow, I am absolutely blown away by your response. Thank you so much for taking the time to share such a detailed, gold-standard breakdown of your experience. It’s clear how much hard work (and analytical pain) went into building this setup!

Your tip about the post-install survey is incredibly smart - it feels like combining metrics from all possible angles is the only way to get a realistic picture. And thanks for explaining the "Referrer Domain" trick, I will definitely check our Data API and weekly exports for those hidden gems.

I have one quick question about your JS redirect bridge: I assume you only use it for external channels like your own blogs, ads, or newsletters? Because from what I know, Atlassian moderators are quite strict and would probably flag or remove such "redirect" links if you posted them inside the Atlassian Community, right?

To answer your question about our stack: we are currently using standard Atlassian Vendor reports (via UI and API) combined with Google Analytics. To be honest, the data still feels super blurry and doesn't give us the full picture yet. Seeing how both you and Natalia look at this, we are also leaning toward treating traffic as a general "touchpoint" bucket and focusing our main analytics on business outcomes like installs and conversions.

Thank you again for your transparency and amazing insights. Feel free to check out Mariem and Natalia's comments in this thread too if you haven't seen them yet - it's turning out to be a really great discussion.

HEMANT SAINI

Hi @Marie -Teamlead-

Thanks so much for the kind words — glad it was useful!

You’re 100% right about the JS redirect bridge. We only use it for external channels we fully control: our blog, docs, ads, newsletters, partner sites, etc.

Inside Atlassian Community we never post the bridge link. Atlassian’s community guidelines are strict about redirecting/shortened links, and mods will remove them fast. For Community posts we just link direct to the Marketplace listing and accept that we’ll lose the UTM. We track Community impact the way you mentioned: date + geo correlation from Vendor reports + “How did you hear about us” survey data. It’s fuzzy, but directionally it works.

One workaround we’ve tested: link to a /resources page on our domain from Community posts instead. That page has helpful docs, comparison guides, and then a clear CTA to the Marketplace. Because the page adds real value beyond just redirecting, it doesn’t get flagged. We still lose last-click attribution, but we capture the assist and time-on-page in GA4 before they head to Marketplace.

Totally aligned with you on treating most traffic as a “touchpoint bucket” and focusing KPIs on installs → trials → conversions. The attribution pain is real, but cohort analysis by week in Looker has been our sanity-saver.

Marie -Teamlead-

Hi Natalia!
Thank you so much for your reply, it’s really valuable.

Thanks again for sharing this, it gives me a lot to think about!

Marie -Teamlead-

Hi Hemant!

Ah, I see what you mean now! That makes perfect sense. By directing people to a valuable / resources page first, you aren't breaking any rules, and you still manage to capture at least that specific portion of the audience that moves from your content to the Marketplace via the CTA link.

Even if it doesn't track 100% of the users (since some might still go and search for the app manually), it’s an amazing way to bridge part of the gap and measure the direct assist from your Community efforts. Definitely a smart workaround!

It's also great to hear that weekly cohort analysis in Looker Studio is what keeps you sane — it really is the only realistic way to look at data without losing your mind over these attribution gaps.

By the way, has the new Marketplace search layout rolled out for you yet? I haven't seen it live on my side still, but I’m really curious to see how it’s going to affect organic traffic and rankings.

Thanks again for being so open and sharing all this "battle-tested" advice. This thread turned out to be way more insightful than I could have hoped for.

Mariem Daghbouji _Caelor_

Hi @Marie -Teamlead- ,

To better clarify my previous point: when we add UTM parameters to a link, we set them up to spell out the exact channel and the exact piece of content. So the source might be "YouTube" and the campaign something like "tabs_tutorial" or "formatting_apps_blog." That way, when someone lands on our listing from that link, we can see in GA4 exactly which video or post sent them, rather than just a vague "referral."

Where things become fuzzy is once users leave that initial tracked path. For example, if someone visits the listing from a blog post, doesn't install immediately, then comes back a few days later through a Marketplace search, a bookmark, or another route, it's very difficult to tie that eventual install back to the original marketing touchpoint. That's the attribution gap we're still trying to solve.

Regarding the Search Keywords report, I really like your idea of looking for keyword trends after major campaigns. I personally see more value in it as a separate source of insight rather than a true attribution tool. For example, if you run a campaign around a specific use case or feature and then notice a corresponding increase in related search terms in the Marketplace, that might suggest the campaign generated awareness and intent. The challenge is that there are usually too many other variables involved (the time of year since people search less over holidays and more at quarter-starts , Atlassian platform events, competitor activity, etc.) to confidently draw a direct connection. So we mainly use it as an additional source of insight for user intent, positioning, messaging, and listing optimization rather than to measure campaign performance.

Still, I do think combining both views can be useful: UTMs tell you what content is driving traffic, while the Search Keywords report helps validate whether your messaging aligns with what users are actively looking for inside the Marketplace.

I'd be really interested if someone has found a way to bridge that gap more reliably, though!

Marie -Teamlead-

Hi @Mariem Daghbouji _Vectors_!

Thanks for the breakdow - the naming convention with explicit sources and campaigns makes total sense.

But I actually meant the underlying technical side of it. Even when we tag everything precisely (similar to your setup), GA4 often tends to flatten or drop that data into generic "direct" or "referral" buckets once it processes the Marketplace listing traffic.

Does this granular tracking work completely out of the box for you without data loss, or did your team have to apply some custom configurations or filters in your GA4 data streams to keep the traffic clean?

Our reports still look incredibly aggregated despite strict UTM practices, so I'm really curious if there's a technical trick or a specific setting we're missing here.

Thanks again, appreciate your insights!

Welcome!

It looks like you're new here. Sign in or register to get started.

Quick Links

Find more posts tagged with

marketplace